SYSTEMS AND METHODS FOR RE-RANKING RANKED SEARCH RESULTS
A system, computer-readable storage medium storing at least one program, and a computer-implemented method for re-ranking ranked search results is presented. Ranked search results satisfying a search query are obtained, where the ranked search results include a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, and where the first search result is ranked higher than the second search result. The first document and the second document are determined to satisfy a similarity criterion. The second entity is determined to satisfy a predefined authorship differential with respect to the first entity. Responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, the second search result and the first search result in the ranked search results are swapped to produce re-ranked search results.
This application is a national stage filing of PCT/US13/41722 (attorney docket No. 060963-5809WO), entitled “Systems and Methods for Re-Ranking Ranked Search Results” filed May 17, 2013, which claims priority to U.S. Provisional Application 61/648,553 (attorney docket No. 060963-5809PR) entitled “Systems and Methods for Re-Ranking Ranked Search Results” filed May 17, 2012, each of which is hereby incorporated by reference herein in its entirety.
This application is also a national stage filing of PCT/US13/41717 (attorney docket No. 060963-5808WO), entitled “Systems and Methods for Crawling and Indexing Content” filed May 17, 2013, which claims priority to U.S. Provisional Application 61/648,546 (attorney docket No. 060963-5808PR) entitled “Systems and Methods for Crawling and Indexing Content,” filed May 17, 2012, each of which is hereby incorporated by reference herein in its entirety.
This application is also a national stage filing of PCT/US13/41723 (attorney docket No. 060963-5810WO), entitled “Systems and Methods for Determining a Likelihood That an Entity is an Author of Original Content” filed May 17, 2013, which claims priority to U.S. Provisional Application 61/648,562 (attorney docket No. 060963-5810PR) entitled “Systems and Methods for Determining a Likelihood That an Entity is an Author of Original Content,” filed May 17, 2012, each of which is hereby incorporated by reference herein in its entirety.
TECHNICAL FIELDThe disclosed embodiments relate generally re-ranking ranked search results.
BACKGROUNDWhen a user submits a search query to a search engine, the search engine identifies search results satisfying the search query and ranks the search results based on a number of ranking factors (e.g., relative relevance of search results). The ranked search results may include two search results that have similar (or substantially the same) content. For example, a first document corresponding to a first search result may be associated with a first entity that is known to not produce original content (e.g., the first entity copies or otherwise redistributes original content authored by other entities). In contrast, a second document corresponding to a second search result may be associated with a second entity that is known to produce original content. In some instances, it is desirable to rank the search results that correspond to documents that are associated with entities that are authors of original content higher than search results corresponding to documents that associated with entities that are not authors of original content even though the documents associated with entities that are not authors of original content have higher relevancy scores.
SUMMARYOne aspect of the present disclosure provides a computer-implemented method for crawling and indexing content, performed on a server system having at least one processor and memory storing at least one program for execution by the at least one processor to perform the method, comprising receiving a message, the message including (i) submitted content, where the submitted content is identified as being published by an entity, and (ii) a link to location on a resource hosting the submitted content; evaluating whether the submitted content is represented in an index of known content to determine whether the submitted content is new relative to the known content; and in response to the evaluating, issuing a request to a search engine to crawl and index the submitted content hosted by the resource associated with the link when the submitted content is new relative to the known content, where the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link is issued responsive to determining that the submitted content is deemed to not be represented in the index of known content.
In some embodiments, the request is a high priority request that is directed to a high priority crawler of the search engine, and where a time when the high priority crawler of the search engine is scheduled to crawl the submitted content occurs before a time when a standard crawler of the search engine is scheduled to crawl the submitted content.
In some embodiments, the evaluating whether the submitted content is represented in the index of known content to determine whether the submitted content is new relative to the known content includes: identifying a set of shingles representing the submitted content, a respective shingle in the set of shingles including a predetermined number of consecutive tokens for the submitted content; comparing the set of shingles representing the submitted content with a first plurality of shingles representing the known content; and when less than a first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, determining that the submitted content is new relative to the known content. In some embodiments, (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, (ii) at least a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, and (iii) at least a third predetermined threshold of corresponding shingles in the second plurality of shingles representing the other submitted content are first instances with respect to corresponding shingles in the first plurality of shingles representing the known content, the method includes: determining that the submitted content is new relative to the known content.
In some embodiments, when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, (ii) at least a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, and (iii) less than a third predetermined threshold of corresponding shingles in the second plurality of shingles representing the other submitted content are first instances with respect to corresponding shingles in the first plurality of shingles representing the known content, the method includes: determining that the submitted content is not new relative to the known content.
In some embodiments when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, and (ii) less than a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, the method includes: determining that the submitted content is not new relative to the known content.
In some embodiments, prior to issuing the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link, the method includes determining that the entity is an author of original content based on an evaluation of other submitted content identified as being published by the entity, where the other submitted content identified as being published by the entity is included in the known content; and the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link is issued responsive to determining that the submitted content is deemed to not be represented in the index of known content and determining that the entity is an author of original content.
In some embodiments, the disclosure further provides determining that the entity is an author of original content based on the other submitted content identified as being published by the entity includes: identifying a second plurality of shingles representing the other submitted content identified as being published by the entity; determining that at least a third predetermined threshold number of the second plurality of shingles representing the other submitted content identified as being published by the entity are first instances with respect to corresponding shingles in a first plurality of shingles representing the known content; and responsive to the determining, determining that the entity is an author of original content.
In some embodiments, the disclosure further provides determining that at least the third predetermined threshold number of the second plurality of shingles representing the other submitted content identified as being published by the entity are first instances with respect to the corresponding shingles in the first plurality of shingles representing the known content includes: determining a first number of the second plurality of shingles representing the other submitted content identified as being published by the entity that are unique with respect to the first plurality of shingles representing the known content; determining a second number of the second plurality of shingles representing the other submitted content identified as being published by the entity that are not unique with respect to the shingles for the known content but that are associated with timestamps that precede timestamps for corresponding shingles in the first plurality of shingles representing the known content; and determining that a sum of the first number and the second number is at least the third predetermined threshold number.
In some embodiments, the disclosure further provides identifying a set of shingles for the submitted content; for each respective shingle in the set of shingles, applying a hash function to the shingle to produce a respective hash value for the respective shingle; and storing the link to a location on the resource hosting the submitted content and a timestamp of the message in a record in the index of known content indexed by the respective hash value.
In some embodiments, for each respective shingle in the set of shingles, storing an identifier for the entity in the record of the index of known content. In some embodiments, the request to the search engine includes the link to the location on the resource and a timestamp associated with the message, and where the timestamp associated with the message is included in a search index entry for the submitted content. In some embodiments, the message is received in response to the entity publishing the submitted content on a website. In some embodiments, the message is received in response to the entity submitting the submitted content and the link to the location on the resource hosting the submitted content using a user interface provided by the server.
In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server, and where the submitted content is marked as being submitted by the entity prior to publication of the submitted content by the entity. In some embodiments, the entity includes at least one author of content. In some embodiments, the entity includes at least one publisher of content. In some embodiments, the entity includes at least one website.
In some embodiments, (i) a respective link to a respective location on a respective resource hosting a respective known content and (ii) a timestamp associated with the respective known content are included in a respective record of the index of known content indexed by a respective hash value for a respective shingle of the respective known content.
In some embodiments, prior to receiving the message, the method includes: receiving, from a client device of the entity, a registration request to register the entity as a registered author of content; providing, to the client device of the entity, at least one verification mechanism to verify that the entity is an author of content; responsive to the providing, receiving, from the client device of the entity, a verification request to verify that the entity is an author of content using the at least one verification mechanism; verifying that the entity is an author of content using the at least one verification mechanism; and registering the entity as a registered author of content.
Another aspect of the present disclosure provides a system to crawl and index content, comprising: at least one processor; memory; and at least one program stored in the memory and executable by the at least one processor, the at least one program comprising instructions to: receive a message, the message including (i) submitted content, where the submitted content is identified as being published by an entity, and (ii) a link to a location on a resource hosting the submitted content; evaluate whether the submitted content is represented in an index of known content to determine whether the submitted content is new relative to the known content; and in response to the evaluating, issue a request to a search engine to crawl and index the submitted content hosted by the resource associated with the link when the submitted content is new relative to the known content, where the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link is issued responsive to determining that the submitted content is deemed to not be represented in the index of known content.
In some embodiments, the request is a high priority request that is directed to a high priority crawler of the search engine, and a time when the high priority crawler of the search engine is scheduled to crawl the submitted content occurs before a time when a standard crawler of the search engine is scheduled to crawl the submitted content.
In some embodiments, the instructions to evaluate whether the submitted content is represented in the index of known content to determine whether the submitted content is new relative to the known content include instructions to: identify a set of shingles representing the submitted content, a respective shingle in the set of shingles including a predetermined number of consecutive tokens for the submitted content; compare the set of shingles representing the submitted content with a first plurality of shingles representing the known content; and when less than a first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, determine that the submitted content is new relative to the known content.
In some embodiments, when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, (ii) at least a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, and (iii) at least a third predetermined threshold of corresponding shingles in the second plurality of shingles representing the other submitted content are first instances with respect to corresponding shingles in the first plurality of shingles representing the known content, the at least one program includes instructions to: determine that the submitted content is new relative to the known content.
In some embodiments, when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, (ii) at least a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, and (iii) less than a third predetermined threshold of corresponding shingles in the second plurality of shingles representing the other submitted content are first instances with respect to corresponding shingles in the first plurality of shingles representing the known content, the at least one program includes instructions to: determine that the submitted content is not new relative to the known content.
In some embodiments, when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, and (ii) less than a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, the at least one program includes instructions to: determine that the submitted content is not new relative to the known content.
In some embodiments, prior to issuing the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link, the at least one program includes instructions to determine that the entity is an author of original content based on an evaluation of other submitted content identified as being published by the entity, where the other submitted content identified as being published by the entity is included in the known content; and where the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link is issued responsive to determining that the submitted content is deemed to not be represented in the index of known content and determining that the entity is an author of original content.
In some embodiments, the instructions to determine that the entity is an author of original content based on the other submitted content identified as being published by the entity include instructions to: identify a second plurality of shingles representing the other submitted content identified as being published by the entity; determine that at least a third predetermined threshold number of the second plurality of shingles representing the other submitted content identified as being published by the entity are first instances with respect to corresponding shingles in a first plurality of shingles representing the known content; and responsive to the determining, determine that the entity is an author of original content.
In some embodiments, the instructions to determine that at least the third predetermined threshold number of the second plurality of shingles representing the other submitted content identified as being published by the entity are first instances with respect to the corresponding shingles in the first plurality of shingles representing the known content include instructions to: determine a first number of the second plurality of shingles representing the other submitted content identified as being published by the entity that are unique with respect to the first plurality of shingles representing the known content; determine a second number of the second plurality of shingles representing the other submitted content identified as being published by the entity that are not unique with respect to the shingles for the known content but that are associated with timestamps that precede timestamps for corresponding shingles in the first plurality of shingles representing the known content; and determine that a sum of the first number and the second number is at least the third predetermined threshold number.
In some embodiments, the computer implemented method further comprises instructions to: identify a set of shingles for the submitted content; for each respective shingle in the set of shingles, apply a hash function to the shingle to produce a respective hash value for the respective shingle; and store the link to the location on the resource hosting the submitted content and a timestamp of the message in a record in the index of known content indexed by the respective hash value.
In some embodiments, the computer implemented method further comprises instructions to, for each respective shingle in the set of shingles, store an identifier for the entity in the record of the index of known content.
In some embodiments, the request to the search engine includes the link to the location on the resource and a timestamp associated with the message, and where the timestamp associated with the message is included in a search index entry for the submitted content.
In some embodiments, the message is received in response to the entity publishing the submitted content on a website. In some embodiments, the message is received in response to the entity submitting the submitted content and the link to the location on the resource hosting the submitted content using a user interface provided by the server.
In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server, and where the submitted content is marked as being submitted by the entity prior to publication of the submitted content by the entity. In some embodiments, the entity includes at least one author of content. In some embodiments, the entity includes at least one publisher of content. In some embodiments, the entity includes at least one website.
In some embodiments, (i) a respective link to a respective location on a respective resource hosting a respective known content and (ii) a timestamp associated with the respective known content are included in a respective record of the index of known content indexed by a respective hash value for a respective shingle of the respective known content. In some embodiments, prior to receiving the message, the at least one program includes instructions to: receive, from a client device of the entity, a registration request to register the entity as a registered author of content; provide, to the client device of the entity, at least one verification mechanism to verify that the entity is an author of content; responsive to the providing, receive, from the client device of the entity, a verification request to verify that the entity is an author of content using the at least one verification mechanism; verify that the entity is an author of content using the at least one verification mechanism; and register the entity as a registered author of content.
Another aspect of the present disclosure provides a non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a computer system, the at least one program comprising instructions to: receive a message, the message including (i) submitted content, where the submitted content is identified as being published by an entity, and (ii) a link to a location on a resource hosting the submitted content; evaluate whether the submitted content is represented in an index of known content to determine whether the submitted content is new relative to the known content; and in response to the evaluating, issue a request to a search engine to crawl and index the submitted content hosted by the resource associated with the link when the submitted content is new relative to the known content, where the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link is issued responsive to determining that the submitted content is deemed to not be represented in the index of known content.
In some embodiments, the request is a high priority request that is directed to a high priority crawler of the search engine, and where a time when the high priority crawler of the search engine is scheduled to crawl the submitted content occurs before a time when a standard crawler of the search engine is scheduled to crawl the submitted content.
In some embodiments, the instructions to evaluate whether the submitted content is represented in the index of known content to determine whether the submitted content is new relative to the known content include instructions to: identify a set of shingles representing the submitted content, a respective shingle in the set of shingles including a predetermined number of consecutive tokens for the submitted content; compare the set of shingles representing the submitted content with a first plurality of shingles representing the known content; and when less than a first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, determine that the submitted content is new relative to the known content.
In some embodiments, when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, (ii) at least a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, and (iii) at least a third predetermined threshold of corresponding shingles in the second plurality of shingles representing the other submitted content are first instances with respect to corresponding shingles in the first plurality of shingles representing the known content, the at least one program includes instructions to: determine that the submitted content is new relative to the known content.
In some embodiments, when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, (ii) at least a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, and (iii) less than a third predetermined threshold of corresponding shingles in the second plurality of shingles representing the other submitted content are first instances with respect to corresponding shingles in the first plurality of shingles representing the known content, the at least one program includes instructions to: determine that the submitted content is not new relative to the known content.
In some embodiments, when (i) at least the first predetermined threshold number of the set of shingles representing the submitted content is included in the first plurality of shingles representing the known content, and (ii) less than a second predetermined threshold number of the set of shingles representing the submitted content is included in a second plurality of shingles representing other submitted content identified as being published by the entity, the at least one program includes instructions to: determine that the submitted content is not new relative to the known content.
In some embodiments, prior to issuing the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link, the at least one program includes instructions to determine that the entity is an author of original content based on an evaluation of other submitted content identified as being published by the entity, where the other submitted content identified as being published by the entity is included in the known content; and where the request to the search engine to crawl and index the submitted content hosted by the resource associated with the link is issued responsive to determining that the submitted content is deemed to not be represented in the index of known content and determining that the entity is an author of original content.
In some embodiments, the instructions to determine that the entity is an author of original content based on the other submitted content identified as being published by the entity include instructions to: identify a second plurality of shingles representing the other submitted content identified as being published by the entity; determine that at least a third predetermined threshold number of the second plurality of shingles representing the other submitted content identified as being published by the entity are first instances with respect to corresponding shingles in a first plurality of shingles representing the known content; and responsive to the determining, determine that the entity is an author of original content.
In some embodiments, the instructions to determine that at least the third predetermined threshold number of the second plurality of shingles representing the other submitted content identified as being published by the entity are first instances with respect to the corresponding shingles in the first plurality of shingles representing the known content include instructions to: determine a first number of the second plurality of shingles representing the other submitted content identified as being published by the entity that are unique with respect to the first plurality of shingles representing the known content; determine a second number of the second plurality of shingles representing the other submitted content identified as being published by the entity that are not unique with respect to the shingles for the known content but that are associated with timestamps that precede timestamps for corresponding shingles in the first plurality of shingles representing the known content; and determine that a sum of the first number and the second number is at least the third predetermined threshold number.
In some embodiments, the non-transitory computer readable storage medium further comprises instructions to: identify a set of shingles for the submitted content; for each respective shingle in the set of shingles, apply a hash function to the shingle to produce a respective hash value for the respective shingle; and store the link to the location on the resource hosting the submitted content and a timestamp of the message in a record in the index of known content indexed by the respective hash value.
In some embodiments, the non-transitory computer readable storage medium further comprises instructions to, for each respective shingle in the set of shingles, store an identifier for the entity in the record of the index of known content. In some embodiments the request to the search engine includes the link to the location on the resource and a timestamp associated with the message, and the timestamp associated with the message is included in a search index entry for the submitted content. In some embodiments, the message is received in response to the entity publishing the submitted content on a website. In some embodiments, the message is received in response to the entity submitting the submitted content and the link to the location on the resource hosting the submitted content using a user interface provided by the server. In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server, and where the submitted content is marked as being submitted by the entity prior to publication of the submitted content by the entity. In some embodiments, the entity includes at least one author of content. In some embodiments, the entity includes at least one publisher of content. In some embodiments, the entity includes at least one website.
In some embodiments, (i) a respective link to a respective location on a respective resource hosting a respective known content and (ii) a timestamp associated with the respective known content are included in a respective record of the index of known content indexed by a respective hash value for a respective shingle of the respective known content.
In some embodiments, prior to receiving the message, the at least one program includes instructions to: receive, from a client device of the entity, a registration request to register the entity as a registered author of content; provide, to the client device of the entity, at least one verification mechanism to verify that the entity is an author of content; responsive to the providing, receive, from the client device of the entity, a verification request to verify that the entity is an author of content using the at least one verification mechanism; verify that the entity is an author of content using the at least one verification mechanism; and register the entity as a registered author of content.
Another aspect of the present disclosure provides a non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a system, the at least one program comprising instructions to be executed by the at least one processor so as to perform any of the methods disclosed herein.
Another aspect of the present disclosure provides a system, comprising at least one processor; and memory storing at least one program for execution by the at least one processor; the system including means for performing the computer-implemented method of any of the methods disclosed herein.
Another aspect of the present disclosure provides a computer-implemented method for re-ranking ranked search results, performed on a server system having at least one processor and memory storing at least one program for execution by the at least one processor to perform the method, comprising: obtaining ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result; determining that the first document and the second document satisfy a similarity criterion; determining that the second entity satisfies a predefined authorship differential with respect to the first entity; and responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, swapping the second search result and the first search result in the ranked search results to produce re-ranked search results.
In some embodiments, the determining that the first document and the second document satisfy the similarity criterion includes: identifying a first plurality of portions of content for the first document; identifying a second plurality of portions of content for the second document; and determining that at least a predetermined number of portions in the first plurality of portions of content and in the second plurality of portions of content are identical. In some embodiments, respective portions of content in the first plurality of portions of content and the second plurality of portions of content include a predetermined amount of data.
In some embodiments, the predetermined number of portions encompasses a predetermined percentage of the data in the first document. In some embodiments, the predetermined number of portions is a threshold percentage of the portions of content in the first plurality of portions of content. In some embodiments, the determining that the first document and the second document satisfy the similarity criterion includes: identifying a first plurality of shingles for the first document; identifying a second plurality of shingles for the second document; and determining that a predetermined quantity of shingles in the first plurality of shingles and in the second plurality of shingles of content are identical.
In some embodiments, the first document and the second document satisfy the similarity criterion when a character count of the first document and a character count of the second document deviate by less than a character count threshold. In some embodiments, the first document and the second document satisfy the similarity criterion when a Levenshtein edit distance from the first document to the second document is less than a predetermined number of character edits.
In some embodiments, the determining that the second entity satisfies the predefined authorship differential with respect to the first entity includes: determining an author score for the first entity and an author score for the second entity, where the author score for a respective entity, in the set of entities consisting of the first entity and the second entity, indicates a likelihood that the respective entity is an author of original content; and when the composite author score for the second entity deviates from the composite author score for the first entity by at least a predetermined threshold, determining that the second entity satisfies the predefined authorship differential with respect to the first entity.
In some embodiments, the determining the author score for a respective entity includes: determining a citation score for the respective entity, where the citation score corresponds to a frequency at which content associated with the respective entity is cited; determining an original author score for the respective entity, where the original author score corresponds to a percentage of content associated with the respective entity that is a first instance of the content in an index of known content; and combining the citation score and the original author score using a predetermined function to produce the author score.
In some embodiments, the determining the citation score for the respective entity includes calculating the citation score as a quotient of a number of citations to content that is associated with the respective entity and a quantity of content that has at least one citation.
In some embodiments, the determining the original author score for the respective entity includes: identifying a plurality of shingles representing content in the index of known content identified as being associated with the respective entity; and calculating a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
In some embodiments the calculating the percentage of the plurality of shingles that are first instances includes: determining a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and calculating the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
In some embodiments, the determining the original author score for the respective entity includes: identifying a plurality of portions of content in the index of known content identified as being associated with the respective entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and calculating a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
In some embodiments, the calculating the percentage of the plurality of portions of content that are first instances includes: determining a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and calculating the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
In some embodiments, when the author score for the first entity and the second entity is within the predetermined threshold of each other, the method includes: obtaining a first timestamp corresponding to a time when the first document was included in an index of known content; obtaining a second timestamp corresponding to a time when the second document was included in the index of known content; and when the second timestamp is less than the first timestamp by at least a predetermined time interval, determining that the second entity satisfies the predefined authorship differential with respect to the first entity.
In some embodiments, the ranked search results are obtained from a search engine, and after swapping the second search result and the first search result in the ranked search results to produce the re-ranked search results, the method further comprises providing the re-ranked search results to the search engine.
Another aspect of the present disclosure provides a system to re-rank ranked search results, comprising: at least one processor; memory; and at least one program stored in the memory and executable by the at least one processor, the at least one program comprising instructions to: obtain ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result; determine that the first document and the second document satisfy a similarity criterion; determine that the second entity satisfies a predefined authorship differential with respect to the first entity; and responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, swap the second search result and the first search result in the ranked search results to produce re-ranked search results.
In some embodiments, the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to: identify a first plurality of portions of content for the first document; identify a second plurality of portions of content for the second document; and determine that at least a predetermined number of portions in the first plurality of portions of content and in the second plurality of portions of content are identical.
In some embodiments, respective portions of content in the first plurality of portions of content and the second plurality of portions of content include a predetermined amount of data. In some embodiments, the predetermined number of portions encompasses a predetermined percentage of the data in the first document. In some embodiments, the predetermined number of portions is a threshold percentage of the portions of content in the first plurality of portions of content. In some embodiments, the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to: identify a first plurality of shingles for the first document; identify a second plurality of shingles for the second document; and determine that a predetermined quantity of shingles in the first plurality of shingles and in the second plurality of shingles of content are identical.
In some embodiments, the first document and the second document satisfy the similarity criterion when a character count of the first document and a character count of the second document deviate by less than a character count threshold. In some embodiments, the first document and the second document satisfy the similarity criterion when a Levenshtein edit distance from the first document to the second document is less than a predetermined number of character edits.
In some embodiments, the instructions to determine that the second entity satisfies the predefined authorship differential with respect to the first entity include instructions to: determine an author score for the first entity and an author score for the second entity, where the author score for a respective entity, in the set of entities consisting of the first entity and the second entity, indicates a likelihood that the respective entity is an author of original content; and when the author score for the second entity is greater than the author score for the first entity by at least a predetermined threshold, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
In some embodiments, the instructions to determine the author score for a respective entity include instructions to: determine a citation score for the respective entity, where the citation score corresponds to a frequency at which content associated with the respective entity is cited; determine an original author score for the respective entity, where the original author score corresponds to a percentage of content associated with the respective entity that is a first instance of the content in an index of known content; and combine the citation score and the original author score using a predetermined function to produce the author score.
In some embodiments, the instructions to determine the citation score for the respective entity include instructions to calculate the citation score as a quotient of a number of citations to content that is associated with the respective entity and a quantity of content that has at least one citation.
In some embodiments, the instructions to determine the original author score for the respective entity include instructions to: identify a plurality of shingles representing content in the index of known content identified as being associated with the respective entity; and calculate a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of shingles that are first instances include instructions to: determine a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
In some embodiments, the instructions to determine the original author score for the respective entity include instructions to: identify a plurality of portions of content in the index of known content identified as being associated with the respective entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and calculate a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of portions of content that are first instances include instructions to: determine a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
In some embodiments, when the author score for the first entity and the second entity is within the predetermined threshold of each other, the at least one program includes instructions to: obtain a first timestamp corresponding to a time when the first document was included in an index of known content; obtain a second timestamp corresponding to a time when the second document was included in the index of known content; and when the second timestamp is less than the first timestamp by at least a predetermined time interval, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
In some embodiments, the ranked search results are obtained from a search engine, and where after swapping the second search result and the first search result in the ranked search results to produce the re-ranked search results, the method further comprises providing the re-ranked search results to the search engine.
Another aspect of the present disclosure provides a non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a computer system, the at least one program comprising instructions to: obtain ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result; determine that the first document and the second document satisfy a similarity criterion; determine that the second entity satisfies a predefined authorship differential with respect to the first entity; and responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, swap the second search result and the first search result in the ranked search results to produce re-ranked search results.
In some embodiments, the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to: identify a first plurality of portions of content for the first document; identify a second plurality of portions of content for the second document; and determine that at least a predetermined number of portions in the first plurality of portions of content and in the second plurality of portions of content are identical.
In some embodiments, respective portions of content in the first plurality of portions of content and the second plurality of portions of content include a predetermined amount of data. In some embodiments, the predetermined number of portions encompasses a predetermined percentage of the data in the first document. In some embodiments, the predetermined number of portions is a threshold percentage of the portions of content in the first plurality of portions of content. In some embodiments, the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to: identify a first plurality of shingles for the first document; identify a second plurality of shingles for the second document; and determine that a predetermined quantity of shingles in the first plurality of shingles and in the second plurality of shingles of content are identical.
In some embodiments, the first document and the second document satisfy the similarity criterion when a character count of the first document and a character count of the second document deviate by less than a character count threshold.
In some embodiments, the first document and the second document satisfy the similarity criterion when a Levenshtein edit distance from the first document to the second document is less than a predetermined number of character edits.
In some embodiments, the instructions to determine that the second entity satisfies the predefined authorship differential with respect to the first entity include instructions to: determine an author score for the first entity and an author score for the second entity, where the author score for a respective entity, in the set of entities consisting of the first entity and the second entity, indicates a likelihood that the respective entity is an author of original content; and when the author score for the second entity is greater than the author score for the first entity by at least a predetermined threshold, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
In some embodiments, the instructions to determine the author score for a respective entity include instructions to: determine a citation score for the respective entity, where the citation score corresponds to a frequency at which content associated with the respective entity is cited; determine an original author score for the respective entity, where the original author score corresponds to a percentage of content associated with the respective entity that is a first instance of the content in an index of known content; and combine the citation score and the original author score using a predetermined function to produce the author score.
In some embodiments, the instructions to determine the citation score for the respective entity include instructions to calculate the citation score as a quotient of a number of citations to content that is associated with the respective entity and a quantity of content that has at least one citation.
In some embodiments, the instructions to determine the original author score for the respective entity include instructions to: identify a plurality of shingles representing content in the index of known content identified as being associated with the respective entity; and calculate a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of shingles that are first instances include instructions to: determine a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
In some embodiments, the instructions to determine the original author score for the respective entity include instructions to: identify a plurality of portions of content in the index of known content identified as being associated with the respective entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and calculate a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of portions of content that are first instances include instructions to: determine a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
In some embodiments, the author score for the first entity and the second entity is within the predetermined threshold of each other, the at least one program includes instructions to: obtain a first timestamp corresponding to a time when the first document was included in an index of known content; obtain a second timestamp corresponding to a time when the second document was included in the index of known content; and when the second timestamp is less than the first timestamp by at least a predetermined time interval, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
In some embodiments, the ranked search results are obtained from a search engine, and where after swapping the second search result and the first search result in the ranked search results to produce the re-ranked search results, the method further comprises providing the re-ranked search results to the search engine.
Another aspect of the present disclosure provides a computer-implemented method for re-ranking ranked search results, performed on a server system having at least one processor and memory storing at least one program for execution by the at least one processor to perform the method, comprising: obtaining ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result; determining that the first document and the second document satisfy a similarity criterion; obtaining a first timestamp corresponding to a time when the first document was included in an index of known content; obtaining a second timestamp corresponding to a time when the second document was included in the index of known content; and swapping the second search result and the first search result in the ranked search results to produce re-ranked search results when the second timestamp is less than the first timestamp by at least a predetermined time interval.
Another aspect of the present disclosure provides a computer-implemented method for determining a likelihood that an entity is an author of original content, performed on a server system having at least one processor and memory storing at least one program for execution by the at least one processor to perform the method, comprising: receiving a message, the message including submitted content, where the submitted content is identified as being associated with an entity; storing a representation of the submitted content in an index of known content; calculating an original author score for the entity based on an evaluation of representations of submitted content in the index of known content identified as being associated with the entity, the original author score indicating a likelihood that the entity is an author of original content; and storing the original author score and an identifier for the entity in an entity database.
In some embodiments, the message comprises a timestamp, and where storing the representation of the submitted content in the index of known content includes: identifying a set of shingles for the submitted content; for each respective shingle in the set of shingles, applying a hash function to the respective shingle to produce a hash value for the respective shingle; and storing the timestamp of the message in a record of the index of known content indexed by the hash value.
In some embodiments, the message includes a link to a location on a resource hosting the submitted content, and where the method further comprises for each respective shingle in the set of shingles, storing the link to the location on the resource hosting the submitted content in a record of the index of known content corresponding to the respective shingle in the set of shingles.
Some embodiments further disclose that, for each respective shingle in the set of shingles, storing the identifier for the entity in a record of the index of known content corresponding to the respective shingle in the set of shingles.
In some embodiments, calculating the original author score for the entity includes: identifying a plurality of shingles representing content, including the submitted content, in the index of known content identified as being associated with the entity; and calculating a percentage of the plurality of shingles that are first instances of the shingles in the index of known content. In some embodiments, calculating the percentage of the plurality of shingles that are first instances includes: determining a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and calculating the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
In some embodiments, calculating the original author score for the entity includes: identifying a plurality of portions of content, including the submitted content, in the index of known content identified as being associated with the entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and calculating a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content. In some embodiments, calculating the percentage of the plurality of portions of content that are first instances includes: determining a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and calculating the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
In some embodiments, the message is received in response to the entity publishing the submitted content on a website. In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server. In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server, and the submitted content is marked as being submitted by the entity prior to publication of the submitted content by the entity. In some embodiments, the entity includes at least one author of content. In some embodiments, the entity includes at least one publisher of content. In some embodiments, the entity includes at least one website.
In some embodiments, prior to receiving the message from the entity, the method includes: receiving, from a client device of the entity, a registration request to register the entity as a registered author of content; providing, to the client device, at least one verification mechanism to verify that the entity is an author of content; responsive to the providing, receiving, from the client device of the entity, a verification request to verify that the entity is an author of content using the at least one verification mechanism; verifying that the entity is an author of content using the at least one verification mechanism; and registering the entity as a registered author of content.
In some embodiments, calculating the original author score for the entity based on the evaluation of representations of the submitted content in the index of known content identified as being associated with the entity includes calculating the original author score for the entity based on the evaluation of a subset of the representations of submitted content in the index of known content identified as being associated with the entity.
In some embodiments, the subset of representations of submitted content includes of all content submitted by the entity within a threshold period of time.
In some embodiments, each representation in the subset of representations of submitted content satisfies a dissimilarity criterion with respect to each other representation in the subset of representations.
In some embodiments, the subset of representations of submitted content consists of representations for ten or less documents submitted by the entity.
Another aspect of the present disclosure provides a system to determine a likelihood that an entity is an author of original content, comprising at least one processor; memory; and at least one program stored in the memory and executable by the at least one processor, the at least one program comprising instructions to: receive a message, the message including submitted content, where the submitted content is identified as being associated with an entity; store a representation of the submitted content in an index of known content; calculate an original author score for the entity based on an evaluation of representations of submitted content in the index of known content identified as being associated with the entity, the original author score indicating a likelihood that the entity is an author of original content; and store the original author score and an identifier for the entity in an entity database.
In some embodiments, the message comprises a timestamp, and where instructions to store the representation of the submitted content in the index of known content include instructions to: identify a set of shingles for the submitted content; for each respective shingle in the set of shingles, apply a hash function to the respective shingle to produce a hash value for the respective shingle; and store the timestamp of the message in a record of the index of known content indexed by the hash value.
In some embodiments, the message includes a link to a location on a resource hosting the submitted content, and where the at least one program includes instructions to, for each respective shingle in the set of shingles, store the link to the location on the resource hosting the submitted content in a record of the index of known content corresponding to the respective shingle in the set of shingles.
In some embodiments, the at least one program includes instructions to, for each respective shingle in the set of shingles, store the identifier for the entity in a record of the index of known content corresponding to the respective shingle in the set of shingles.
In some embodiments, the instructions to calculate the original author score for the entity include instructions to: identify a plurality of shingles representing content, including the submitted content, in the index of known content identified as being associated with the entity; and calculate a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of shingles that are first instances include instructions to: determine a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
In some embodiments, the instructions to calculate the original author score for the entity include instructions to: identify a plurality of portions of content, including the submitted content, in the index of known content identified as being associated with the entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and calculate a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of portions of content that are first instances include instructions to: determine a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
In some embodiments, the message is received in response to the entity publishing the submitted content on a website. In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server. In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server, and where the submitted content is marked as being submitted by the entity prior to publication of the submitted content by the entity.
In some embodiments, the entity includes at least one author of content. In some embodiments, the entity includes at least one publisher of content. In some embodiments, the entity includes at least one website.
In some embodiments, the at least one program includes instructions to, prior to receiving the message from the entity: receive, from a client device of the entity, a registration request to register the entity as a registered author of content; provide, to the client device, at least one verification mechanism to verify that the entity is an author of content; responsive to the providing, receive, from the client device of the entity, a verification request to verify that the entity is an author of content using the at least one verification mechanism; verify that the entity is an author of content using the at least one verification mechanism; and register the entity as a registered author of content.
In some embodiments, the instructions to calculate the original author score for the entity based on the evaluation of representations of the submitted content in the index of known content identified as being associated with the entity include instructions to calculate the original author score for the entity based on the evaluation of a subset of the representations of submitted content in the index of known content identified as being associated with the entity.
In some embodiments, the subset of representations of submitted content includes of all content submitted by the entity within a threshold period of time.
In some embodiments, each representation in the subset of representations of submitted content satisfies a dissimilarity criterion with respect to each other representation in the subset of representations.
In some embodiments, the subset of representations of submitted content consists of representations for ten or less documents submitted by the entity.
Another aspect of the present disclosure provides a non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a computer system, the at least one program comprising instructions to: receive a message, the message including submitted content, where the submitted content is identified as being associated with an entity; store a representation of the submitted content in an index of known content; calculate an original author score for the entity based on an evaluation of representations of submitted content in the index of known content identified as being associated with the entity, the original author score indicating a likelihood that the entity is an author of original content; and store the original author score and an identifier for the entity in an entity database.
In some embodiments, the message comprises a timestamp, and the instructions to store the representation of the submitted content in the index of known content include instructions to: identify a set of shingles for the submitted content; for each respective shingle in the set of shingles, apply a hash function to the respective shingle to produce a hash value for the respective shingle; and store the timestamp of the message in a record of the index of known content indexed by the hash value.
In some embodiments, the message includes a link to a location on a resource hosting the submitted content, and where the at least one program includes instructions to, for each respective shingle in the set of shingles, store the link to the location on the resource hosting the submitted content in a record of the index of known content corresponding to the respective shingle in the set of shingles.
In some embodiments, the at least one program includes instructions to, for each respective shingle in the set of shingles, store the identifier for the entity in a record of the index of known content corresponding to the respective shingle in the set of shingles.
In some embodiments, the instructions to calculate the original author score for the entity include instructions to: identify a plurality of shingles representing content, including the submitted content, in the index of known content identified as being associated with the entity; and calculate a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of shingles that are first instances include instructions to: determine a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
In some embodiments, the instructions to calculate the original author score for the entity include instructions to: identify a plurality of portions of content, including the submitted content, in the index of known content identified as being associated with the entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and calculate a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
In some embodiments, the instructions to calculate the percentage of the plurality of portions of content that are first instances include instructions to: determine a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and calculate the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
In some embodiments, the message is received in response to the entity publishing the submitted content on a website. In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server.
In some embodiments, the message is received in response to the entity submitting the submitted content using a user interface provided by the server, and where the submitted content is marked as being submitted by the entity prior to publication of the submitted content by the entity. In some embodiments, the entity includes at least one author of content. In some embodiments, the entity includes at least one publisher of content. In some embodiments, where the entity includes at least one website.
In some embodiments, the at least one program includes instructions to, prior to receiving the message from the entity: receive, from a client device of the entity, a registration request to register the entity as a registered author of content; provide, to the client device, at least one verification mechanism to verify that the entity is an author of content; responsive to the providing, receive, from the client device of the entity, a verification request to verify that the entity is an author of content using the at least one verification mechanism; verify that the entity is an author of content using the at least one verification mechanism; and register the entity as a registered author of content.
In some embodiments, the instructions to calculate the original author score for the entity based on the evaluation of representations of the submitted content in the index of known content identified as being associated with the entity include instructions to calculate the original author score for the entity based on the evaluation of a subset of the representations of submitted content in the index of known content identified as being associated with the entity.
In some embodiments, the subset of representations of submitted content consists of all content submitted by the entity within a threshold period of time.
In some embodiments, each representation in the subset of representations of submitted content satisfies a dissimilarity criterion with respect to each other representation in the subset of representations.
In some embodiments, the subset of representations of submitted content consists of representations for ten or less documents submitted by the entity.
The embodiments disclosed herein are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings. Like reference numerals refer to corresponding parts throughout the drawings.
The embodiments described herein provide techniques for re-ranking ranked search results.
In some embodiments, a respective resource server 102 is associated with a respective entity 105. For example, in some embodiments the respective entity 105 is an operator of the respective resource server 102. In other embodiments, the respective entity 105 is a subscriber of a publication service (e.g., a news wire service) hosted by the respective resource server 102, and/or a subscriber of a web hosting service of the respective resource server 102.
In some implementations, the respective entity 105 includes at least one author of content. For example, the respective entity 105 may include an individual author or one of a plurality of co-authors for (or contributors to) content. In some embodiments, an entity 105 is a business organization that produces original, or partially original content. In some embodiments, an entity 105 is a news organization. In some implementations, the entity 105 includes at least one publisher of content. For example, the respective entity 105 may be a publisher of books, a publisher of periodicals, a publisher of online content and/or the like. In some implementations, the respective entity 105 is the author of content on at least one website. For example, the respective entity 105 may contribute original content to a blogging website, a website for a publisher (e.g., news, magazine, etc.) and/or the like. Note that such a website may include a subset of the content within a particular domain. For example, the website may include content in a particular domain (e.g., a top-level domain example.com). In another example, the website includes content in a sub-domain of the particular domain (e.g., a sub-domain blogs.example.com). In another example, the website includes content in a directory of the domain (e.g., www.example.com/˜johndoe/). In some embodiments, the website includes content in: a plurality of domains (e.g., a network of affiliated websites), a plurality of sub-domains of at least one domain, and/or a plurality of subdirectories of at least one domain. In some embodiments, the content authored by an entity 105 is a blog post, a social network post, or a post in an on-line discussion thread. In some embodiments, the content authored by the entity 105 is any content that has been posted to a location accessible on the Internet such that it is readily ascertainable that the entity 105 posted the content.
Thus, network system 100 encompasses a broad array of systems 100. For example, in some embodiments (i) a plurality of entities 105 author or publish content (e.g., books, periodicals, websites, etc.) that is hosted on a single resource server 102, (ii) a plurality of entities 105 author content that is stored on a plurality of resource servers 102, and/or (iii) a single entity 105 authors content that is stored on a plurality of resource servers 102. Any combination of such arrangements is also found in the network systems 100 within the scope of the present disclosure. For instance, an example network system 100 provides for one entity 105 that authors content that is stored on two resource servers 102 while two or more other entities 105 author content that is stored on a single resource server 102.
The search engine 112 periodically crawls links to identify and index new (or previously un-indexed) content hosted by resource servers 102. The frequency at which the search engine 112 crawls links for a particular domain may depend on factors including, but not limited to, a frequency at which new (or previously un-indexed content) is available on the particular domain, an amount of traffic the domain receives from users (e.g., entities), an importance of web pages in the particular domain, whether the domain includes original content, and the like. The search engine 112 also identifies ranked search results that satisfy search queries received from client devices 104.
The client devices 104 include any computing device that the entities 105 may use to access the resource servers 102 (e.g., a website or content hosted on the resource servers 102, etc.), the server 110, and/or the search engine 112. For example, the client devices 104 may include a laptop computer, a desktop computer, a smartphone, a mobile phone, a personal digital assistant, a tablet device, a set top box, a game console, and/or the like.
In some embodiments, the entity 105-1 uses the client device 104-1 to publish content onto the resource server 102-1. For example, the resource server 102-1 may host content for a newspaper website and the entity 105-1 may use the client device 104-1 to upload and to publish a news article onto the newspaper website hosted on the resource server 102-1. Note that the entity 105-1 may upload the news article to the resource server 102-1 without publishing the news article. For example, the entity 105-1 may upload the news article to the resource server 102-1 and publish the news article at a later date and/or time. The act of publishing content makes the content available to users (e.g., to the public, to subscribers, etc.).
As discussed above, one problem faced by search engines is that content uploaded to the resource server 102-1 may not be indexed and may not be available to the search engine until after the usefulness of the content has passed. In order to address this problem, in some embodiments, the resource server 102-1 sends a message to the server 110 to notify the server 110 that the content has been published on the resource server 102-1. Alternatively, in some embodiments, the entity 105-1 uses the client device 104-1 to send a message to the server 110 to notify the server 110 that the content has been published on the resource server 102-1. For example, the entity 105-1 may use the client device 104-1 to send a message to the server 110 using a web page provided by the server 110 that allows the entity 105-1 to notify the server 110 that the content has been published on the resource server 102-1. In either of these alternative embodiments, the message includes all or a portion of the content, termed the submitted content 231, which was published on the resource server 102-1 and/or a link to the published content on the resource server 102-1. The term “submitted content” is used in this specification to refer to the content included in the message or the content that is a target of a link that is included in the message. The server 110 may then conditionally issue a request to the search engine 112 to crawl and index the content (i.e., the submitted content 231) that was uploaded to the resource server 102-1. In some embodiments, the request is a high-priority request that is directed to a high priority crawler of the search engine 112. In these embodiments, a time when the high priority crawler of the search engine 112 is scheduled to crawl the content occurs before a time when a standard crawler of the search engine 112 is scheduled to crawl the submitted content 231. These embodiments are described in more detail below with reference to
Some embodiments of the disclosed systems and methods further address the problem incurred by crawling and indexing when faced with duplicate or substantially duplicate existing content. Such duplicate content poses a challenge to the identification of authors of original content so that such original content may be attributed with authorship of content and accordingly returned in response to user queries. Accordingly, in some embodiments, the server 110 determines a likelihood that an entity is an author of original content. These embodiments are described in more detail below with reference to
Note that although
Also note that although the embodiments described herein refer to the server 110 and the search engine 112, the embodiments may be applied to multiple servers and search engines. Furthermore, the functionality of any of the server 110 and the search engine 112 may be implemented within a single server (or a set of distributed servers). For example, the server 110 and the search engine 112 may be located on the same server (or the same set of distributed servers).
In an alternative embodiment not depicted in
In some embodiments, the message 230 includes all of the content that was published on the resource server 102-1 and a link to the content that was published on the resource server 102-1. In some embodiments, the message 230 includes a portion of the content that was published on the resource server 102-1 and a link to the content that was published on the resource server 102-1. For example, the message 230 may contain the first kilobyte, first 10 kilobytes, first 100 kilobytes or first megabyte of the content that was published on the resource server 102-1. In some embodiments, the message 230 includes a digest of the content that was published on the resource server 102-1 and a link to the content that was published on the resource server 102-1. In some embodiments, the message 230 only includes a link to the content that was published on the resource server 102-1. In these embodiments, the server 110 obtains all or a portion of the content published on the resource server 102-1 prior to proceeding with the other operations described below. For example, in some embodiments the server 110 obtains the first kilobyte, first 10 kilobytes, first 100 kilobytes or first megabyte of the content that was published on the resource server 102-1. In another example, in some embodiments the server 110 obtains a digest of the content that was published on the resource server 102-1. This digest may be the result of, for example, a predetermined hash function on the known content. As discussed above, the term “submitted content” is used in this specification to refer to the content included in the message 230 or the content that is a target of a link that is included in the message 230.
Returning to the embodiment illustrated in
In some embodiments, the index of known content 220 includes content obtained through crawl operations of the search engine 112. In some embodiments, the index of known content 220 includes a representation of content obtained through messages (e.g., the message 230) received from the resource server 102-1 (or the client device 104-1). The content is obtained either directly from a message 230 when the message 230 includes the submitted content 231 or indirectly when the message 230 includes a link to the submitted content 231.
As illustrated in
In some embodiments, the content module 204 stores a representation of the submitted content 231 in the index of known content 220. For example, in some instances, the content module 204 generates a set of shingles for the submitted content 231, and for each respective shingle in the set of shingles, the content module 204 (i) applies a hash function to the shingle to produce a corresponding hash value for the respective shingle and (ii) stores information relating to the submitted content 231 (e.g., a link to a location on the resource server 102-1 hosting the submitted content 231, a timestamp of the message 230, etc.) in a record in the index of known content 220 that is indexed by the respective hash value.
The embodiments described in
Returning to
In some embodiments, the message 330 only includes a link to the content that was published on the resource server 102-1. In some embodiments, the message 330 includes a portion of the content that was published on the resource server 102-1 and a link to the content that was published on the resource server 102-1. For example, the message 330 may contain the first kilobyte, first 10 kilobytes, first 100 kilobytes or first megabyte of the content that was published on the resource server 102-1. In some embodiments, the message 330 may contain a digest of the content that was published on the resource server 102-1. In embodiments where the message 330 does not include any portion of the content but does contain a link to the content, the server 110 obtains the content published on the resource server 102-1 prior to proceeding with the other operations described below. In some embodiments where the message 330 includes a portion of the content but not all of the content, the server 110 obtains the full set of content published on the resource server 102-1 prior to proceeding with the other operations described below. In some alternative embodiments where the message 330 includes a portion of the content but not all of the content, the server 110 does not obtain the full set of content published on the resource server 102-1 prior to proceeding with the other operations described below.
The front end module 202 receives the message 330 and provides the message 330 to the content module 204. The content module 204 stores a representation of the submitted content 331 in the index of known content 220. In some embodiments, the content module 204 stores a representation of the submitted content 331 in the index of known content 220. For example, in some embodiments, the content module 204 generates a set of shingles for the submitted content 331 and, for each respective shingle in the set of shingles, the content module 204 (i) applies a hash function to the respective shingle to produce a corresponding hash value for the respective shingle and (ii) stores information relating to the submitted content 331 (e.g., a link to a location on the resource server 102-1 hosting the submitted content 331, a timestamp of the message 330, etc.) in a record in the index of known content 220 that is indexed by the respective hash value.
As illustrated in
In some embodiments, the author module 206 calculates the original author score for the entity 105-1 in response to receiving the message 330. In some embodiments, the author module 206 periodically calculates the original author score for the entity 105-1 (e.g., every hour, every day, every week, etc.). In some embodiments, all content associated with the entity 105-1 in the index of known content 220 is considered in computing the original author score for the entity 105-1. In some embodiments, only more recent content associated with the entity 105-1 is considered in computing the original author score for the entity 105-1. For example, in various embodiments, only content within a threshold time period is used to calculate the original author score of the entity 105-1 (e.g., content with a timestamp within the past week, the past month, the past year, or the past two years, etc.). In some embodiments, only a discrete set of content associated with the entity 105-1 is considered in computing the original author score for the entity 105-1. For example, in various embodiments, only five or less, ten or less, fifty or less, or 100 or less different documents identified as being associated with the entity 105-1 are used to calculate the original author score of the entity 105-1. In some embodiments, recent content associated with the entity 105-1 in the index of known content provides a more significant contribution to the original author score for the entity 105-1 than less recent content. To illustrate one such embodiment, in some original authorship calculations, more recent content is up-weighted and less recent content is down-weighted in the original author score calculations.
In some embodiments, only content that is unique with respect to all other content that (i) has already been selected to compute an original author score for the entity 105-1 and (ii) is associated with the entity 105-1 is considered when computing the original author score for the entity. For example, in some embodiments, only content that (i) satisfies a dissimilarity criterion (e.g., at least 80 percent different, at least 90 percent different, etc.) with respect to all other content that has been selected for computation of an original author score for the entity 105-1 and (ii) is associated with the entity 105-1 is used in computing the original author score for the entity 105-1. In this way, original author scores are not biased by multiple submissions of the original same original work by the entity 105-1.
Dissimilarity between a first document and second document associated with the entity 105-1 can be computed using any measure of similarity or dissimilarity in order to determine with the dissimilarity criterion has been satisfied. In some embodiments, a first document and second document associated with the entity 105-1 satisfy the dissimilarity criterion when a Levenshtein edit distance from the first document to the second document is greater than a predetermined number of character edits. In some embodiments, the first and second documents satisfy the dissimilarity criterion when a Jaro-Winkler distance metric from the first document to the second document is more than a predetermined distance. In some embodiments, the first and second documents satisfy the dissimilarity criterion when a Jaccard index is less than a predetermined value. In some embodiments, the first and second documents satisfy the dissimilarity criterion when a Masi distance is greater than a predetermined distance.
In some embodiments, the first and second documents document satisfy the dissimilarity criterion when a character count of the first document and a character count of the second document deviate by more than a character count threshold (and/or a word count threshold). For example, if the character count threshold is five characters, the first document and the second document satisfy the dissimilarity criterion when the character count of the first document and the character count of the second document deviate by more than five characters.
In some embodiments, the first and second documents document satisfy the dissimilarity criterion when the first document contains a predetermined quantity of shingles that are not found in the second document, or vice versa, where such shingles are computed using the techniques disclosed below in conjunction with step 1202 of
In some embodiments, the author module 206 stores the original author score 332 in a record of an author database 222 that is associated with the entity 105-1.
The embodiments described in
When the ranked search results 430 includes a first search result corresponding to a first document associated with a first entity (e.g., entity 105-1) and a second search result corresponding to a second document associated with a second entity (e.g., entity 105-2) where (1) the first search result is ranked higher than the second search result, (2) the first document and the second document satisfy a similarity criterion (e.g., based on information obtained from the index of known content 220), and (3) the second entity satisfies a predefined authorship differential with respect to the first entity (e.g., based on original author scores obtained from the author database 222), the re-ranking module 208 swaps the first ranked search result and the second ranked search result to produce re-ranked search results 431.
As illustrated in
The embodiments described in
In accordance with the example embodiments illustrated in
As illustrated in
In an alternative to the embodiment depicted in
-
- an operating system 612 that includes procedures for handling various basic system services and for performing hardware dependent tasks;
- a communication module 614 that is used for connecting the server 110 to other computers via the one or more communication interfaces 604 (wired or wireless) and one or more communication networks, such as the Internet, other wide area networks, local area networks, metropolitan area networks, and so on;
- an optional user interface module 616 that receives commands from the user via the input devices 608 and generates user interface objects in the display device 606;
- the front end module 202 that provides an interface between the server 110 and other computer systems, as described herein;
- the content module 204 that evaluates submitted content to determine whether the submitted content is new relative to known content in the index of known content 220, issues requests to the search engine 112 to crawl and index the submitted content, and/or stores a representation of the submitted content in the index of known content 220, as described herein;
- the author module 206 that calculates an original author score for an entity based at least in part on an evaluation of representations of submitted content in the index of known content 220 identified as being associated with the entity, as described herein;
- the re-ranking module 208 that re-ranks ranked search results based at least in part on an original author score for the entities associated with the ranked search results, as described herein;
- the registration module 210 that registers an entity as a registered author of content, as described herein;
- the index of known content 220 that includes representations of submitted content, as described herein; and
- the author database 222 that includes entity information for entities (e.g., the entity information 533).
In some embodiments, the programs or modules identified above correspond to sets of instructions for performing a function described above. The sets of instructions can be executed by one or more processors (e.g., the CPUs 602). The above identified modules or programs (i.e., sets of instructions) need not be implemented as separate software programs, procedures or modules, and thus various subsets of these programs or modules may be combined or otherwise re-arranged in various embodiments. In some embodiments, memory 610 stores a subset of the modules and data structures identified above. Furthermore, memory 610 may store additional modules and data structures not described above.
Although
-
- an operating system 712 that includes procedures for handling various basic system services and for performing hardware dependent tasks;
- a communication module 714 that is used for connecting the resource server 102-1 to other computers via the one or more communication interfaces 704 (wired or wireless) and one or more communication networks, such as the Internet, other wide area networks, local area networks, metropolitan area networks, and so on;
- an optional user interface module 716 that receives commands from the user via the input devices 708 and generates user interface objects in the display device 706;
- a web module 718 that responds to requests to access web pages or other resources (e.g., scripts, programs, etc.) hosted on the resource server 102-1; and
- a publication module 720 that publishes content for entities and sends messages (e.g., the messages 230 and 330) to the server 110 indicating that content has been published on the resource server 102-1.
In some embodiments, the programs or modules identified above correspond to sets of instructions for performing a function described above. The sets of instructions can be executed by one or more processors (e.g., the CPUs 702). The above identified modules or programs (i.e., sets of instructions) need not be implemented as separate software programs, procedures or modules, and thus various subsets of these programs or modules may be combined or otherwise re-arranged in various embodiments. In some embodiments, memory 710 stores a subset of the modules and data structures identified above. Furthermore, memory 710 may store additional modules and data structures not described above.
Although
-
- an operating system 812 that includes procedures for handling various basic system services and for performing hardware dependent tasks;
- a communication module 814 that is used for connecting the search engine 112 to other computers via the one or more communication interfaces 804 (wired or wireless) and one or more communication networks, such as the Internet, other wide area networks, local area networks, metropolitan area networks, and so on;
- an optional user interface module 816 that receives commands from the user via the input devices 808 and generates user interface objects in the display device 806;
- a search module 818 that identifies ranked search results that satisfy a search query;
- a crawler module 820 that periodically performs normal-priority crawl operations on links to identify new content and performs high-priority crawl operations to identify new content in response to receiving a request from the content module 204 of the server 110; and
- an indexer module 822 that indexes new content identified by the crawler module 820 into a search index 824.
In some embodiments, the programs or modules identified above correspond to sets of instructions for performing a function described above. The sets of instructions can be executed by one or more processors (e.g., the CPUs 802). The above identified modules or programs (i.e., sets of instructions) need not be implemented as separate software programs, procedures or modules, and thus various subsets of these programs or modules may be combined or otherwise re-arranged in various embodiments. In some embodiments, memory 810 stores a subset of the modules and data structures identified above. Furthermore, memory 810 may store additional modules and data structures not described above.
Although
-
- an operating system 912 that includes procedures for handling various basic system services and for performing hardware dependent tasks;
- a communication module 914 that is used for connecting the client device 104-1 to other computers via the one or more communication interfaces 904 (wired or wireless) and one or more communication networks, such as the Internet, other wide area networks, local area networks, metropolitan area networks, and so on;
- a user interface module 916 that receives commands from the user via the input devices 908 and generates user interface objects in the display device 906; and
- a browser module 918 that provides a user interface for users to access and obtain documents and/or services hosted on servers (e.g., the resource servers 102), renders the documents and/or services in the user interface, and/or executes scripts (or other code).
In some embodiments, the programs or modules identified above correspond to sets of instructions for performing a function described above. The sets of instructions can be executed by one or more processors (e.g., the CPUs 902). The above identified modules or programs (i.e., sets of instructions) need not be implemented as separate software programs, procedures or modules, and thus various subsets of these programs or modules may be combined or otherwise re-arranged in various embodiments. In some embodiments, memory 910 stores a subset of the modules and data structures identified above. Furthermore, memory 910 may store additional modules and data structures not described above.
Although
Re-Ranking Ranked Search Results
The embodiments described below with reference to
The following discussion refers to the resource server 102-1, the client device 104-1, the entity 105-1, and the search engine 112. However, it should be noted that the following discussion may be applied to any resource server, client device, entity, and search engine. Furthermore, the following discussion refers to particular modules of the server 110 performing particular operations illustrated in the
The re-ranking module 208 determines (1004) that the first document and the second document satisfy a similarity criterion. In some embodiments, the first document and the second document satisfy the similarity criterion when a similarity threshold has been exceeded. For example, if the similarity criterion is based on the percentage of the content in the first and second documents that are similar, the similarity threshold may be set to a value of 90%. Accordingly, when the percentage of the content in the first and second documents that are similar exceeds the 90% similarity threshold, the re-ranking module 208 determines that the first document and the second document satisfy the similarity criterion. Operation 1004 is described in more detail below with reference to
The re-ranking module 208 determines (1006) that the second entity satisfies a predefined authorship differential with respect to the first entity. In some embodiments, the authorship differential is a difference between the original author scores for the first and second entities. Operation 1006 is described in more detail below with reference to
Responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, the re-ranking module 208 swaps (1008) the second search result and the first search result in the ranked search results to produce re-ranked search results.
In some embodiments, when the ranked search results are obtained from the search engine 112, after swapping the second search result and the first search result in the ranked search results to produce the re-ranked search results, the re-ranking module 208 provides the re-ranked search results to the search engine 112.
In some embodiments, rather than swapping the second search result and the first search result in the ranked search results to produce re-ranked search results as specified in (1008), the second search result is given a higher ranking that is not necessarily that of the first search result. For example, in some implementations, the second search result is given the ranking of the first result and the first result is deleted from the ranked search results. In another example, the first search result is given a ranking in the ranked search results that is less than its original ranking and the second search result is given a ranking that is greater than its original ranking in the ranked search results. In this way, the second search result obtains a ranking that is higher than the ranking of the first search result in the revised ranked search results. In yet another example, the second search result is given a ranking that is greater than it original ranking such that its ranking is higher than the ranking of the first search result in the revised ranked search results.
In some embodiments, respective portions of content in the first plurality of portions of content and the second plurality of portions of content include a predetermined amount of data. For example, the predetermined amount of data may include a predetermined number of characters, a predetermined number of bytes, a predetermined number of words, and the like.
In some embodiments, the shingles in a respective plurality of shingles include at least one overlapping token (e.g., word or character). For example, if the content is “a big dog crosses the street” and 3-shingling is used (i.e., the set of all unique combinations of three words in the content is used as the basis for shingling), the set of shingles includes the following shingles: “a big dog,” “big dog crosses,” “dog crosses the,” and “crosses the street.” In some embodiments, the shingles in the respective plurality of shingles do not include overlapping tokens. For example, using the example content from the previous example, the set of shingles for a 3-shingling of the content includes the following shingles: “a big dog,” and “crosses the street.” Note that the discussion with reference to
In some embodiments, the first document and the second document satisfy the similarity criterion when a character count of the first document and a character count of the second document deviate by less than a character count threshold (and/or a word count threshold). For example, if the character count threshold is five characters, the first document and the second document satisfy the similarity criterion when the character count of the first document and the character count of the second document deviate by less than five characters.
In some embodiments, the first document and the second document satisfy the similarity criterion when a Levenshtein edit distance from the first document to the second document is less than a predetermined number of character edits. In some embodiments, the first document and the second document satisfy the similarity criterion when a Jaro-Winkler distance metric from the first document to the second document is less than a predetermined distance. In some embodiments, the first document and the second document satisfy the similarity criterion when a Jaccard index is greater than a predetermined value. In some embodiments, the first document and the second document satisfy the similarity criterion when a Masi distance is less than a predetermined distance.
When the author score for the second entity is greater than the author score for the first entity by at least a predetermined threshold (e.g., 10%), the re-ranking module 208 determines (1304) that the second entity satisfies the predefined authorship differential with respect to the first entity.
The re-ranking module 208 determines (1404) an original author score for the respective entity, where the original author score corresponds to a percentage of content associated with the respective entity that is a first instance of the content in the index of known content 220. Operation 1404 is described in more detail below with reference to
The re-ranking module 208 combines (1406) the citation score and the original author score using a predetermined function (e.g., simple summation of the citation score and the original author score, a weighted sum of the citation score and the original author score, multiplication of the citation score and the original author score, etc.) to produce the author score.
In some embodiments, the re-ranking module 208 determines a first number of the plurality of shingles that are unique with respect to all other shingles in the index of known content and determines a second number of the plurality of shingles that are not unique with respect to the corresponding shingles in the index of known content but that are associated with timestamps that precede timestamps for identical shingles in the index of known content. The re-ranking module 208 then calculates the percentage as a quotient of (i) a sum of the first number and the second number and (ii) a total number of shingles in the plurality of shingles.
In some embodiments, the re-ranking module 208 calculates (1704) a percentage of portions of content that are first instances by determining a first number of the plurality of portions of content that are unique with respect to all other portions of content in the index of known content and determining a second number of the plurality of portions of content that are not unique with respect to the corresponding portions of content in the index of known content but that are associated with timestamps that precede timestamps for identical portions of content in the index of known content. The re-ranking module 208 then calculates the percentage as a quotient of (i) a sum of the first number and the second number and (ii) a total number of portions of content in the plurality of portions of content.
In some embodiments, the re-ranking module 208 determines (1404) an original author score for a respective entity by identifying an amount of content in the index of known content identified as being associated with the entity, where the amount of content includes the submitted content. The re-ranking module 208 then calculates a percentage of the amount that is uniquely represented in the index of known content. In some implementations, the re-ranking module 208 calculates the percentage of the amount by determining a subset of the amount for which there are no identical portions of content in the index of known content having an earlier timestamp and calculating the percentage as a quotient of the subset and the amount.
In some embodiments, when the second entity does not satisfy a predefined authorship differential with respect to the first entity (e.g., the other branch of operation 1006), the re-ranking module 208 compares timestamps associated with the first document and the second document to determine whether to swap the first and second search results.
The methods illustrated in
Plural instances may be provided for components, operations or structures described herein as a single instance. Finally, boundaries between various components, operations, and data stores are somewhat arbitrary, and particular operations are illustrated in the context of specific illustrative configurations. Other allocations of functionality are envisioned and may fall within the scope of the implementation(s). In general, structures and functionality presented as separate components in the example configurations may be implemented as a combined structure or component. Similarly, structures and functionality presented as a single component may be implemented as separate components. These and other variations, modifications, additions, and improvements fall within the scope of the implementation(s).
It will also be understood that, although the terms “first,” “second,” etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first contact could be termed a second contact, and, similarly, a second contact could be termed a first contact, which changing the meaning of the description, so long as all occurrences of the “first contact” are renamed consistently and all occurrences of the second contact are renamed consistently. The first contact and the second contact are both contacts, but they are not the same contact.
The terminology used herein is for the purpose of describing particular implementations only and is not intended to be limiting of the claims. As used in the description of the implementations and the appended claims, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that the term “and/or” as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items. It will be further understood that the terms “comprises” and/or “comprising,” when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
As used herein, the term “if” may be construed to mean “when” or “upon” or “in response to determining” or “in accordance with a determination” or “in response to detecting,” that a stated condition precedent is true, depending on the context. Similarly, the phrase “if it is determined (that a stated condition precedent is true)” or “if (a stated condition precedent is true)” or “when (a stated condition precedent is true)” may be construed to mean “upon determining” or “in response to determining” or “in accordance with a determination” or “upon detecting” or “in response to detecting” that the stated condition precedent is true, depending on the context.
Embodiments in which a second search result and a first search result are swapped in ranked search results have been presented above. In an alternative implementation of any of these embodiments, rather than swapping the second search result and the first search result in the ranked search results to produce re-ranked search results as specified, the second search result is given a higher ranking that is not necessarily that of the first search result. For example, in some embodiments, the second search result is given the ranking of the first result and the first result is deleted from the ranked search results. In another example, the first search result is given a ranking in the ranked search results that is less than its original ranking and the second search result is given a ranking that is greater than its original ranking in the ranked search results. In this way, the second search result obtains a ranking that is higher than the ranking of the first search result in the revised ranked search results. In yet another example, the second search result is given a ranking that is greater than it original ranking such that its ranking is higher than the ranking of the first search result in the revised ranked search results.
The foregoing description included example systems, methods, techniques, instruction sequences, and computing machine program products that embody illustrative implementations. For purposes of explanation, numerous specific details were set forth in order to provide an understanding of various implementations of the inventive subject matter. It will be evident, however, to those skilled in the art that implementations of the inventive subject matter may be practiced without these specific details. In general, well-known instruction instances, protocols, structures and techniques have not been shown in detail.
The foregoing description, for purpose of explanation, has been described with reference to specific implementations. However, the illustrative discussions above are not intended to be exhaustive or to limit the implementations to the precise forms disclosed. Many modifications and variations are possible in view of the above teachings. The implementations were chosen and described in order to best explain the principles and their practical applications, to thereby enable others skilled in the art to best utilize the implementations and various implementations with various modifications as are suited to the particular use contemplated.
Claims
1. A computer-implemented method for re-ranking ranked search results, performed on a server system having at least one processor and memory storing at least one program for execution by the at least one processor to perform the method, comprising:
- obtaining ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result;
- determining that the first document and the second document satisfy a similarity criterion;
- determining that the second entity satisfies a predefined authorship differential with respect to the first entity; and
- responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, swapping the second search result and the first search result in the ranked search results to produce re-ranked search results.
2. The computer-implemented method of claim 1, wherein determining that the first document and the second document satisfy the similarity criterion includes:
- identifying a first plurality of portions of content for the first document;
- identifying a second plurality of portions of content for the second document; and
- determining that at least a predetermined number of portions in the first plurality of portions of content and in the second plurality of portions of content are identical.
3. The computer-implemented method of claim 2, wherein respective portions of content in the first plurality of portions of content and the second plurality of portions of content include a predetermined amount of data.
4. The computer-implemented method of claim 2, wherein the predetermined number of portions encompasses a predetermined percentage of the data in the first document.
5. The computer-implemented method of claim 2, wherein the predetermined number of portions is a threshold percentage of the portions of content in the first plurality of portions of content.
6. The computer-implemented method of any of claims 1-5, wherein determining that the first document and the second document satisfy the similarity criterion includes:
- identifying a first plurality of shingles for the first document;
- identifying a second plurality of shingles for the second document; and
- determining that a predetermined quantity of shingles in the first plurality of shingles and in the second plurality of shingles of content are identical.
7. The computer-implemented method of any of claims 1-5, wherein the first document and the second document satisfy the similarity criterion when a character count of the first document and a character count of the second document deviate by less than a character count threshold.
8. The computer-implemented method of any of claims 1-5, wherein the first document and the second document satisfy the similarity criterion when a Levenshtein edit distance from the first document to the second document is less than a predetermined number of character edits.
9. The computer-implemented method of any of claims 1-8, wherein determining that the second entity satisfies the predefined authorship differential with respect to the first entity includes:
- determining an author score for the first entity and an author score for the second entity, wherein the author score for a respective entity, in the set of entities consisting of the first entity and the second entity, indicates a likelihood that the respective entity is an author of original content; and
- when the composite author score for the second entity deviates from the composite author score for the first entity by at least a predetermined threshold, determining that the second entity satisfies the predefined authorship differential with respect to the first entity.
10. The computer-implemented method of claim 9, wherein determining the author score for a respective entity includes:
- determining a citation score for the respective entity, wherein the citation score corresponds to a frequency at which content associated with the respective entity is cited;
- determining an original author score for the respective entity, wherein the original author score corresponds to a percentage of content associated with the respective entity that is a first instance of the content in an index of known content; and
- combining the citation score and the original author score using a predetermined function to produce the author score.
11. The computer-implemented method of claim 10, wherein determining the citation score for the respective entity includes calculating the citation score as a quotient of a number of citations to content that is associated with the respective entity and a quantity of content that has at least one citation.
12. The computer-implemented method of any of claims 10-11, wherein determining the original author score for the respective entity includes:
- identifying a plurality of shingles representing content in the index of known content identified as being associated with the respective entity; and
- calculating a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
13. The computer-implemented method of claim 12, wherein calculating the percentage of the plurality of shingles that are first instances includes:
- determining a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and
- calculating the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
14. The computer-implemented method of any of claims 10-11, wherein determining the original author score for the respective entity includes:
- identifying a plurality of portions of content in the index of known content identified as being associated with the respective entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and
- calculating a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
15. The computer-implemented method of claim 14, wherein calculating the percentage of the plurality of portions of content that are first instances includes:
- determining a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and
- calculating the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
16. The computer-implemented method of claim 9, wherein, when the author score for the first entity and the second entity is within the predetermined threshold of each other, the method includes:
- obtaining a first timestamp corresponding to a time when the first document was included in an index of known content;
- obtaining a second timestamp corresponding to a time when the second document was included in the index of known content; and
- when the second timestamp is less than the first timestamp by at least a predetermined time interval, determining that the second entity satisfies the predefined authorship differential with respect to the first entity.
17. The computer-implemented method of claim 1, wherein the ranked search results are obtained from a search engine, and wherein after swapping the second search result and the first search result in the ranked search results to produce the re-ranked search results, the method further comprises providing the re-ranked search results to the search engine.
18. A system to re-rank ranked search results, comprising:
- at least one processor;
- memory; and
- at least one program stored in the memory and executable by the at least one processor, the at least one program comprising instructions to: obtain ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result; determine that the first document and the second document satisfy a similarity criterion; determine that the second entity satisfies a predefined authorship differential with respect to the first entity; and responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, swap the second search result and the first search result in the ranked search results to produce re-ranked search results.
19. The system of claim 18, wherein the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to:
- identify a first plurality of portions of content for the first document;
- identify a second plurality of portions of content for the second document; and
- determine that at least a predetermined number of portions in the first plurality of portions of content and in the second plurality of portions of content are identical.
20. The system of claim 19, wherein respective portions of content in the first plurality of portions of content and the second plurality of portions of content include a predetermined amount of data.
21. The system of claim 19, wherein the predetermined number of portions encompasses a predetermined percentage of the data in the first document.
22. The system of claim 19, wherein the predetermined number of portions is a threshold percentage of the portions of content in the first plurality of portions of content.
23. The system of any of claims 18-22, wherein the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to:
- identify a first plurality of shingles for the first document;
- identify a second plurality of shingles for the second document; and
- determine that a predetermined quantity of shingles in the first plurality of shingles and in the second plurality of shingles of content are identical.
24. The system of any of claims 18-22, wherein the first document and the second document satisfy the similarity criterion when a character count of the first document and a character count of the second document deviate by less than a character count threshold.
25. The system of any of claims 18-22, wherein the first document and the second document satisfy the similarity criterion when a Levenshtein edit distance from the first document to the second document is less than a predetermined number of character edits.
26. The system of any of claims 18-25, wherein the instructions to determine that the second entity satisfies the predefined authorship differential with respect to the first entity include instructions to:
- determine an author score for the first entity and an author score for the second entity, wherein the author score for a respective entity, in the set of entities consisting of the first entity and the second entity, indicates a likelihood that the respective entity is an author of original content; and
- when the author score for the second entity is greater than the author score for the first entity by at least a predetermined threshold, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
27. The system of claim 26, wherein the instructions to determine the author score for a respective entity include instructions to:
- determine a citation score for the respective entity, wherein the citation score corresponds to a frequency at which content associated with the respective entity is cited;
- determine an original author score for the respective entity, wherein the original author score corresponds to a percentage of content associated with the respective entity that is a first instance of the content in an index of known content; and
- combine the citation score and the original author score using a predetermined function to produce the author score.
28. The system of claim 27, wherein the instructions to determine the citation score for the respective entity include instructions to calculate the citation score as a quotient of a number of citations to content that is associated with the respective entity and a quantity of content that has at least one citation.
29. The system of any of claims 27-28, wherein the instructions to determine the original author score for the respective entity include instructions to:
- identify a plurality of shingles representing content in the index of known content identified as being associated with the respective entity; and
- calculate a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
30. The system of claim 29, wherein the instructions to calculate the percentage of the plurality of shingles that are first instances include instructions to:
- determine a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and
- calculate the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
31. The system of any of claims 27-28, wherein the instructions to determine the original author score for the respective entity include instructions to:
- identify a plurality of portions of content in the index of known content identified as being associated with the respective entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and
- calculate a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
32. The system of claim 31, wherein the instructions to calculate the percentage of the plurality of portions of content that are first instances include instructions to:
- determine a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and
- calculate the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
33. The system of claim 26, wherein, when the author score for the first entity and the second entity is within the predetermined threshold of each other, the at least one program includes instructions to:
- obtain a first timestamp corresponding to a time when the first document was included in an index of known content;
- obtain a second timestamp corresponding to a time when the second document was included in the index of known content; and
- when the second timestamp is less than the first timestamp by at least a predetermined time interval, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
34. The system of claim 18, wherein the ranked search results are obtained from a search engine, and wherein after swapping the second search result and the first search result in the ranked search results to produce the re-ranked search results, the method further comprises providing the re-ranked search results to the search engine.
35. A non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a computer system, the at least one program comprising instructions to:
- obtain ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result;
- determine that the first document and the second document satisfy a similarity criterion;
- determine that the second entity satisfies a predefined authorship differential with respect to the first entity; and
- responsive to determining that the second entity satisfies the predefined authorship differential with respect to the first entity, swap the second search result and the first search result in the ranked search results to produce re-ranked search results.
36. The non-transitory computer readable storage medium of claim 35, wherein the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to:
- identify a first plurality of portions of content for the first document;
- identify a second plurality of portions of content for the second document; and
- determine that at least a predetermined number of portions in the first plurality of portions of content and in the second plurality of portions of content are identical.
37. The non-transitory computer readable storage medium of claim 36, wherein respective portions of content in the first plurality of portions of content and the second plurality of portions of content include a predetermined amount of data.
38. The non-transitory computer readable storage medium of claim 36, wherein the predetermined number of portions encompasses a predetermined percentage of the data in the first document.
39. The non-transitory computer readable storage medium of claim 36, wherein the predetermined number of portions is a threshold percentage of the portions of content in the first plurality of portions of content.
40. The non-transitory computer readable storage medium of any of claims 35-39, wherein the instructions to determine that the first document and the second document satisfy the similarity criterion include instructions to:
- identify a first plurality of shingles for the first document;
- identify a second plurality of shingles for the second document; and
- determine that a predetermined quantity of shingles in the first plurality of shingles and in the second plurality of shingles of content are identical.
41. The non-transitory computer readable storage medium of any of claims 35-39, wherein the first document and the second document satisfy the similarity criterion when a character count of the first document and a character count of the second document deviate by less than a character count threshold.
42. The non-transitory computer readable storage medium of any of claims 35-39, wherein the first document and the second document satisfy the similarity criterion when a Levenshtein edit distance from the first document to the second document is less than a predetermined number of character edits.
43. The non-transitory computer readable storage medium of any of claims 35-42, wherein the instructions to determine that the second entity satisfies the predefined authorship differential with respect to the first entity include instructions to:
- determine an author score for the first entity and an author score for the second entity, wherein the author score for a respective entity, in the set of entities consisting of the first entity and the second entity, indicates a likelihood that the respective entity is an author of original content; and
- when the author score for the second entity is greater than the author score for the first entity by at least a predetermined threshold, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
44. The non-transitory computer readable storage medium of claim 43, wherein the instructions to determine the author score for a respective entity include instructions to:
- determine a citation score for the respective entity, wherein the citation score corresponds to a frequency at which content associated with the respective entity is cited;
- determine an original author score for the respective entity, wherein the original author score corresponds to a percentage of content associated with the respective entity that is a first instance of the content in an index of known content; and
- combine the citation score and the original author score using a predetermined function to produce the author score.
45. The non-transitory computer readable storage medium of claim 44, wherein the instructions to determine the citation score for the respective entity include instructions to calculate the citation score as a quotient of a number of citations to content that is associated with the respective entity and a quantity of content that has at least one citation.
46. The non-transitory computer readable storage medium of any of claims 44-45, wherein the instructions to determine the original author score for the respective entity include instructions to:
- identify a plurality of shingles representing content in the index of known content identified as being associated with the respective entity; and
- calculate a percentage of the plurality of shingles that are first instances of the shingles in the index of known content.
47. The non-transitory computer readable storage medium of claim 46, wherein the instructions to calculate the percentage of the plurality of shingles that are first instances include instructions to:
- determine a first number of the plurality of shingles, each having an associated timestamp, for which there are no identical shingles in the index of known content having an earlier timestamp; and
- calculate the percentage as a quotient of the first number and a total number of shingles in the plurality of shingles.
48. The non-transitory computer readable storage medium of any of claims 44-45, wherein the instructions to determine the original author score for the respective entity include instructions to:
- identify a plurality of portions of content in the index of known content identified as being associated with the respective entity, each portion in the plurality of portions representing a predetermined amount of data in the index of known content; and
- calculate a percentage of the plurality of the portions that are first instances of the portions of content in the index of known content.
49. The non-transitory computer readable storage medium of claim 48, wherein the instructions to calculate the percentage of the plurality of portions of content that are first instances include instructions to:
- determine a first number of the plurality of portions of content, each having an associated timestamp, for which there are no identical portions of content in the index of known content having an earlier timestamp; and
- calculate the percentage as a quotient of the first number and a total number of portions of content in the plurality of portions of content.
50. The non-transitory computer readable storage medium of claim 43, wherein, when the author score for the first entity and the second entity is within the predetermined threshold of each other, the at least one program includes instructions to:
- obtain a first timestamp corresponding to a time when the first document was included in an index of known content;
- obtain a second timestamp corresponding to a time when the second document was included in the index of known content; and
- when the second timestamp is less than the first timestamp by at least a predetermined time interval, determine that the second entity satisfies the predefined authorship differential with respect to the first entity.
51. The non-transitory computer readable storage medium of claim 35, wherein the ranked search results are obtained from a search engine, and wherein after swapping the second search result and the first search result in the ranked search results to produce the re-ranked search results, the method further comprises providing the re-ranked search results to the search engine.
52. A non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a system, the at least one program comprising instructions to be executed by the at least one processor so as to perform the method of any of claims 1-17.
53. A system, comprising:
- at least one processor; and
- memory storing at least one program for execution by the at least one processor;
- the system including means for performing the method of any of claims 1-17.
54. A computer-implemented method for re-ranking ranked search results, performed on a server system having at least one processor and memory storing at least one program for execution by the at least one processor to perform the method, comprising:
- obtaining ranked search results satisfying a search query, the ranked search results including a first search result corresponding to a first document associated with a first entity and a second search result corresponding to a second document associated with a second entity, the first search result being ranked higher than the second search result;
- determining that the first document and the second document satisfy a similarity criterion;
- obtaining a first timestamp corresponding to a time when the first document was included in an index of known content;
- obtaining a second timestamp corresponding to a time when the second document was included in the index of known content; and
- swapping the second search result and the first search result in the ranked search results to produce re-ranked search results when the second timestamp is less than the first timestamp by at least a predetermined time interval.
55. A non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a system, the at least one program comprising instructions to be executed by the at least one processor so as to perform the method of claim 54.
56. A system, comprising:
- at least one processor; and
- memory storing at least one program for execution by the at least one processor;
- the system including means for performing the method of claim 54.
Type: Application
Filed: May 17, 2013
Publication Date: Jun 18, 2015
Inventors: Chung Tin Kwok (Mountain View, CA), Lei Zhong (Sunnyvale, CA), Zhihuan Qiu (Sunnyvale, CA)
Application Number: 14/401,828
