Abstract: Methods and systems are provided for creating an adaptive thesaurus. A term pair including an index term and an expansion term is received. A recall gain, an expansion independence, and a semantic similarity of the term pair are calculated using a processor. Whether to store the term pair is determined based on the recall gain, the expansion independence, and the semantic similarity. The term pair is stored based on the determination. Methods and systems are provided for searching using an adaptive thesaurus. A search query including a query term is received. An expansion term stored in association with an index term matching the query term in the adaptive thesaurus is retrieved. Using a processor, the search query is expanded using the expansion term based on a recall gain, an expansion independence, and a semantic similarity.

Abstract: A technique for finding the nearest match in a computer storage system is provided. A query statement includes a new operator that indicates that a user desires to access a set of rows that contain a value nearest to a target value. An index is accessed that is based at least in part on a column reference included in the statement. The index comprises a plurality of leaf nodes, where each leaf node comprises one or more entries, where each entry contains a key value, corresponding to the column reference, and a reference to a row in a table. Because leaf nodes in an index are ordered and linked to one another, a portion of the index need only be scanned once. The set of rows from the table are returned that are referenced by one or more entries whose column values are nearest to the target value.

Abstract: An improved duplicate detection technique that uses query-relevant information to limit the portion(s) of documents to be compared for similarity is described. Before comparing two documents for similarity, the content of these documents may be condensed based on the query. In one embodiment, query-relevant information or text (also referred to as “snippets”) is extracted from the documents and only the extracted snippets, rather than the entire documents, are compared for purposes of determining similarity.

Abstract: Systems and methods that improve search rankings for a search query by using data associated with queries related to the search query are described. In one aspect, a search query is received, a related query related to the search query is determined, an article (such as a web page) associated with the search query is determined, and a ranking score for the article based at least in part on data associated with the related query is determined. Several algorithms and types of data associated with related queries useful in carrying out such systems and methods are described.

Abstract: A database query optimizer includes an access plan storage mechanism that stores execution detail corresponding to a query statement in an access plan cache. By storing execution detail in the access plan cache along with the query statement, the execution detail is available when needed to debug database problems, even after a problem occurs. The execution detail stored in the access plan cache may be mined and analyzed after-the-fact, producing a very powerful tool for debugging database problems. Snapshots of the execution detail may also be created by a user. In addition, the user may setup autonomic snapshots and comparisons so that a user may be alerted from a degradation in database performance between two snapshots.

Abstract: Database query analysis technology, in which an input query is received that relates to multiple attribute classes of data records. An index that is descriptive of the data records in data storage is accessed and relative positions in the accessed index for the multiple attribute classes are identified. The accessed index is processed based on more than one of the multiple attribute classes and the identified relative positions, and the processed index is used to identify groups of data records that are included in the data records in the data storage and that share common values for the multiple attribute classes in the input query. Metrics corresponding to the data records included in each of the identified groups of data records are accessed, computations are performed on the accessed metrics, and a report is generated that reflects results of the computations.

Abstract: Overall latency is reduced when processing tasks such as search queries by determining which tasks are “expensive,” or likely to exceed desired latency thresholds. For expensive queries processed according to partitions, the segments for each partition can be divided among various sub-queries, which allow each partition to be processed in parallel by multiple devices without the need for repartitioning. Further, the responses to the sub-queries can be monitored, and if one or more responses are not received within a specified amount of time then each sub-query for which a response is missing can be resent. The first response received will be consolidated with the results from the other queries, and the result returned.

Abstract: A computer-implemented method is provided, including receiving a search query from a user during a search session, and presenting information to the user responsively to the search query. After an indication that the user takes an action related to the search session is received, a portion of the presented information is identified that the user did not select prior to taking the action, and respective scores are assigned to search results generated responsively to the search query, including lowering one or more of the scores of the respective search results that are characterized by the non-selected portion of the presented information. The search results are ranked according to the scores, and presented to the user. Other embodiments are also described.

Abstract: Methods, systems, apparatus, including computer program products, for providing fresh related search suggestions in response to a user submitted query are presented. In one implementation, a plurality of prior queries are selected wherein each of the prior queries was submitted as a search query a number of times during a recent time period and satisfies a criterion. For each of the prior queries, the prior is selected as a candidate query based on one or more of: a determination that search results responsive to the prior query include a number of news results that satisfy a second threshold, and relevance data indicative of user behavior relative to the search results responsive to the prior query. In response to receiving a user query, one or more candidate queries are selected that match the user query.

Abstract: A method of searching for and downloading a mobile telephone file is disclosed, and in the method, a server in a domain where a mobile terminal is located receives a keyword in a searching request, searches for resources in the mobile communication network, and returns a list of resources searched out to the mobile terminal; the mobile terminal receives data resource information selected by a user according to the list of the resources the mobile terminal sends a request for downloading resources to the server in the domain, and the server sends the resources to the mobile terminal. An apparatus of searching for and downloading a mobile telephone file is provided also, which includes a storage module, a receiving module, a sending module, a query module, and can also include a data base, a communication interface and a statistic module. A system of searching for and downloading a mobile telephone file is provided.

Abstract: A search server system having one or more processors and memory receives a first search query from a client and, in response, generates a first set of search results corresponding to the first search query by searching in a first set of one or more index portions and generates a query cursor indicating a location in the first set of index portions. The server sends the first set of search results and the query cursor to the client. Subsequently the server receives, from the client, a second search query including the query cursor. The first search query is different from the second search query and/or different index portions are used to perform the second search query. In response to the second search query the server generates a second set of search results identified using the query cursor and sends the second set of search results to the client.

Abstract: A search server system having one or more processors and memory receives, from a client, a current search query including a query cursor corresponding to a prior search query. The query cursor corresponds to a location in a first set of index portions used to perform the prior search query. In response to the current search query, the server decomposes the query cursor into a plurality of components and generates a translated query cursor corresponding to the received search query using one or more of the components. The server subsequently identifies a location in a second set of index portions to start performing the current search query based the translated query cursor. The server generates a set of search results corresponding to the current search query by searching in the second set of index portions starting at the identified location and sends search results to the client.

Abstract: Systems and methods that improve search rankings for a search query by using data associated with queries related to the search query are described. In one aspect, a search query is received, a related query related to the search query is determined, an article (such as a web page) associated with the search query is determined, and a ranking score for the article based at least in part on data associated with the related query is determined. Several algorithms and types of data associated with related queries useful in carrying out such systems and methods are described.

Abstract: Suggested search results are provided with suggested search queries as a user incrementally enters characters of a search query. At each incremental user input, a query portion is received and suggested search queries are identified based on the query portion. Suggested search results associated with the suggested search queries are also identified. These suggested search results enable direct navigation to web sites associated with the search results. In some embodiments, a user's search history is analyzed to identify search results selected by the user for the suggested search queries. In other embodiments, search histories for a group of users associated with the user are analyzed to identify the suggested search results to provide with the suggested search queries.

Abstract: Improved approaches to completing, submitting and updating on-line forms or registrations are disclosed. In one embodiment, an on-line form completion agent or an on-line registration agent intelligently guides an on-line form completion or registration process. In another embodiment, a registration is automatically monitored to determine whether and when the registration needs updated.

Abstract: Methods, systems, and apparatus, including computer program products, for ranking search results of a search query using corpus search statistics. In one aspect, a method includes determining a first relevance of a first corpus to a search query, determining a second relevance of a second corpus to the search query, determining a measure of relative relevance of the first corpus and the second corpus to the search query, and providing the measure of relative relevance to a ranking engine for ranking of search results for a new search corresponding to the search query.

Abstract: A domain-specific sentiment classifier that can be used to score the polarity and magnitude of sentiment expressed by domain-specific documents is created. A domain-independent sentiment lexicon is established and a classifier uses the lexicon to score sentiment of domain-specific documents. Sets of high-sentiment documents having positive and negative polarities are identified. The n-grams within the high-sentiment documents are filtered to remove extremely common n-grams. The filtered n-grams are saved as a domain-specific sentiment lexicon and are used as features in a model. The model is trained using a set of training documents which may be manually or automatically labeled as to their overall sentiment to produce sentiment scores for the n-grams in the domain-specific sentiment lexicon. This lexicon is used by the domain-specific sentiment classifier.

Abstract: According to some embodiments, supplemental data may be retrieved from a supplemental data source. It may be automatically detected that a first personal insurance information value in the supplemental data source is correlated with a first personal insurance information value in the original insurance claim records source. A second personal insurance information value that was previously used to replace the first personal insurance information value in the original insurance claim records source may then be retrieved, and the first personal insurance information value in the supplemental data source may be automatically replaced with that same second personal insurance information value to provide referential integrity.

Abstract: Methods and apparatus consistent with the invention provide improved organization of documents responsive to a search query. In one embodiment, a search query is received and a list of responsive documents is identified. The responsive documents are organized based in whole or in part on usage statistics.

Abstract: A database monitor tracks performance statistics and information about the execution of different SQL statements. A query optimizer benefits from these statistics when generating an access plan. In particular, the query optimizer, upon receiving an SQL statement, searches the records of the database monitor for similar SQL statements that have previously been executed. As part of determining the best access plan for the current SQL statement, the query optimizer considers the information retrieved from the database monitor. In this way, the access plan that is generated can automatically be tuned based on empirical performance evidence.

Abstract: Approaches, techniques, and mechanisms are disclosed for maintaining a history of query plans executed for a database command, along with information related to each query plan. A database server receives a request to execute a particular command. The database server determines a plan for executing the particular command. The database server adds first information to a plan history associated with the particular command. The plan history comprises information related to a plurality of plans that have been generated for the particular command. The first information may include, for example, properties of the plan (including an outline of the plan) as well as statistics collected during execution of the plan. The database server may implement techniques for periodically refreshing information in a plan history. The database server may also implement techniques for purging old or less important plans.

Abstract: One embodiment of the present invention provides a system that automatically identifies compounds, such as bigrams or n-grams. During operation, the system obtains selections of search results which were selected by one or more users, wherein the search results were previously generated by a search engine in response to queries containing search terms. Next, the system forms a set of candidate compounds from the queries, wherein each candidate compound comprises n consecutive terms from a query. Then, for each candidate compound in the set, the system analyzes the selections of search results to calculate a likelihood that the candidate compound is a compound.

Abstract: A system ranks results. The system may receive a list of links. The system may identify a source with which each of the links is associated and rank the list of links based at least in part on a quality of the identified sources.

Abstract: A journaling subsystem on a mobile device stores event data related to applications or other subsystems running on the mobile device. The event data can be stored and indexed in a journal database so that a timeline of past events can be reconstructed in response to search queries. In some implementations, a timeline can be reconstructed with markers on a map display based on search results. When the user interacts with a marker on the map display, the event data collected by the mobile device is made available to the user.

Abstract: Embodiments of the invention provide techniques for optimizing database queries for energy efficiency. In general, a query optimizer is configured to compare energy requirements of query plans, and to select a query plan requiring minimal energy to execute. In one embodiment, the query optimizer may also compare time performance of the query plans, and may select a query plan by matching to a user preference for a relative priority between energy requirements and time performance.

Abstract: A search query is received. Personal information for a user is then determined. A search is performed in a general subdomain of general content using the search query. For example, the general subdomain of general content may be a WWW search. Then, a vertical subdomain is determined based on the personal information. A search is then performed in the vertical subdomain of specialized content using the search query. The search performed in the general subdomain and the search performed in the vertical subdomain generate general search results and vertical search results. The results may be combined and outputted to a client.

Abstract: A technique for information retrieval includes parsing a corpus to identify a number of wordform instances within each document of the corpus. A weighted morpheme-by-document matrix is generated based at least in part on the number of wordform instances within each document of the corpus and based at least in part on a weighting function. The weighted morpheme-by-document matrix separately enumerates instances of stems and affixes. Additionally or alternatively, a term-by-term alignment matrix may be generated based at least in part on the number of wordform instances within each document of the corpus. At least one lower rank approximation matrix is generated by factorizing the weighted morpheme-by-document matrix and/or the term-by-term alignment matrix.

Abstract: Document management techniques to account for user-specific patterns in document metadata are disclosed. In one embodiment, a method for facilitating document retrieval may comprise: assigning a first entitlement to a first user for accessing a first plurality of documents; identifying patterns in the first user's creation or modification of metadata related to the first plurality of documents; recording the identified patterns associated with the first user; receiving a document query from a second user who has been assigned a second entitlement to access a second plurality of documents; determining, based on the second entitlement, an access right of the second user with respect to the first plurality of documents; and modifying the document query based on the access right of the second user and the identified patterns, such that the document query returns relevant documents from the first plurality of documents despite the second user's ignorance of the identified patterns.

Abstract: A type of search engine (referred to as the “Get Engine”) receives one or more keywords, semantically formulates a question being asked from the keywords, generates specifications for the query, and searches a website index to determine websites that are likely owners of the answer to the question based on the query specifications and website classifications. The Get Engine determines a website that is most likely the owner of the answer based on credibility, searches the pages of the website using the keywords and additional keywords related to the query, retrieves the answer from the pages of the website, and receives feedback used in part to determine the credibility of the website.

Abstract: In one embodiment, a method includes concurrently executing a set of multiple queries, through a processor, to improve a resource usage within a data warehouse system. The method also includes permitting a group of users of the data warehouse system to simultaneously run a set of queries. In addition, the method includes applying a high-concurrency query operator to continuously optimize a large number of concurrent queries for a set of highly concurrent dynamic workloads.

Abstract: A method and system for providing up-to-date monitoring results in a virtual machine environment are described. In one embodiment, a host controller receives a new query pertaining to at least one object in a virtual machine system from a client, and determines whether the new query is similar to any queries previously received from other clients. If the new query is similar to any of the previously received queries, the host controller provides data pertaining to a result of a last execution of the similar query to the client. If the new query is not similar to any of the previously received queries, the host controller executes the new query and provides a result of an execution of the new query to the client.

Abstract: Systems and methods which facilitate search engine discovery of and/or access to application program content are shown. Declarative code may be provided within the procedural code of a rich Internet application in order to identify various states, and their associated content, of the rich Internet application. A Web crawler may identify content associated with a rich Internet application using the foregoing information. Likewise, direct access to rich Internet application states associated with the identified content may be provided using the foregoing information. A translation module may be provided for interfacing between a Web crawler and a rich Internet application.

Abstract: One embodiment is a method that generates points from an input space obtained from a query for a database. A kernel function maps the points from the input space to a feature space. Given a point in the feature space, a reverse mapping identifies coordinates in the input space for the point to predict performance of the query before the query executes in the database.

Abstract: A computer-implemented method includes receiving information expressing a user's interest in one or more media programs, obtaining information indicative of popularity for a plurality of media programs responsive to the received information by individuals other than the user, and transmitting one or more recommendations of media programs for display to the user, from the plurality of media programs that relate to the received information.

Abstract: A dynamic data buffer method, includes providing a system for accessing a database having a plurality of fields of data potentially used by applications, analyzing an application's structure to identify at least one field of data from the plurality of fields of data in the database used by the application, performing a field-wise select for selecting only the identified at least one field of data and creating a dynamic data buffer for storing only the identified at least one field of data.

Abstract: A search server system having one or more processors and memory receives, from a client, a current search query including a query cursor corresponding to a prior search query. The query cursor corresponds to a location in a first set of index portions used to perform the prior search query. In response to the current search query, the server decomposes the query cursor into a plurality of components and generates a translated query cursor corresponding to the received search query using one or more of the components. The server subsequently identifies a location in a second set of index portions to start performing the current search query based the translated query cursor. The server generates a set of search results corresponding to the current search query by searching in the second set of index portions starting at the identified location and sends search results to the client.

Abstract: A computer-implemented method includes calculating first statistics about a user-identified event within a first subset of a database of events; selecting a second subset of the database of events based on said first statistics; calculating second statistics about the user-identified event within the second subset of the database of events; merging the first and second statistics as statistics of the user-identified event within the entire database of events; and generating a result including at least a portion of the merged statistics of the user-identified event.

Abstract: Disclosed is a method and device and computer storage medium for determining whether two pages linked on the World Wide Web are a part of the same World Wide Web site. The method involves examining and comparing the IP addresses of the Web pages, and can also be extended to finding other pages to which a given Web page is linked on the Web, and to determining whether a Web page of interest is part of a Web site with a desired characteristic, such as being part of an electronic commerce site.

Abstract: A method is disclosed. The method includes receiving a message from a remote device. The message includes data. The method further includes analyzing the data using a set of rules. The analyzing includes performing a maintenance analysis by analyzing the data using a set of asset maintenance rules and a set of asset usage rules. Responsive to the analyzing the data using the set of rules, an alert is sent.

Abstract: A computer method and apparatus of social tagging computer resources is disclosed. In the invention method and apparatus, a data store holds social tagging data of previously followed tags throughout a system. In response to end-user indication of a tag candidate, the invention system retrieves from the data store social tagging data about the tag candidate. A tag previewer forms and renders a display of the retrieved social tagging data. The display indicates number of users having interest in (i.e., having followed, searched with, browsed and/or subscribed to) the candidate tag and its variants, and the type of interest. The display may further indicate identities (e.g., email addresses) of these users having such interest. The display may serve as a precommit preview to the end-user during selection of a tag (the candidate) for a given resource. Alternatively, the tag previewer may provide the invention display of social tagging data during end-user browsing as well.

Abstract: In a networked computer system that includes multiple nodes and multiple networks interconnecting the nodes, a database query optimizer takes network choice into consideration when optimizing a query. The query optimizer logs historical information for queries that includes network information. A network monitor determines network characteristics of the multiple networks. The query optimizer then selects a network for a query based on the network characteristics, and optionally based on the historical information as well. The result is a query optimizer that takes network choice into consideration, thereby providing a query that is optimized according to the characteristics of the networks.

Abstract: Methods and apparatus, including computer program products, are provided that include techniques for searching a database. One method includes receiving one or more input search terms and providing a database. The database maps at least one search term to at least one coupon. The method further includes identifying one or more coupons in the database based on the received input search terms. The method further includes selecting one or more coupons from the identified coupons. A given coupon can be selected based on a number of times the given coupon has been previously printed. A given coupon can also be selected based on a number of times the given coupon has been previously redeemed. The method further includes presenting one or more coupons to the user where the presented one or more coupons include the selected one or more coupons.

Abstract: The present disclosure includes a system and method for identifying negative keywords associated with advertisements. In some implementations, a method for advertising includes identifying one or more keywords associated with an advertisement. The keywords used to embed the advertisement in Web pages such as those including search results. By scoring keywords to identify those keywords that are not relevant to the web page content, one or more negative keywords can be identified in order to limit the embedding of non-relevant advertisements in web page content.

Abstract: Techniques for determining search query information for an advertising campaign and communicating the search query information to a sponsor of the advertising campaign are described. The techniques include receiving, from a sponsor of the advertisement, a request for search query information relating to the advertisement having been presented by a search service, identifying previously executed search queries submitted by users of the search service, wherein search results of the identified search queries have been presented with the advertisement to one or more users of the search service, compiling search query information, the search query information including ad performance data associated with the advertisement for the identified search queries, and providing the sponsor with the compiled search query information. A user interface may be generated to receive user input of selection criteria and provide an interface to search query information.

Abstract: A computer-implemented method for controlling content distribution includes forwarding information associated with a user to a device operated by the user, the information being configured for use in selecting content from any of multiple content providers for a content distribution to the user. The method includes receiving, in response to the information, an edit of the information forwarded from the device. The edit identifies a first content provider and including a first modification of the content distribution regarding the first content provider. The method includes storing the edit in association with the information such that the first modification is taken into account in the content distribution. The method can be implemented using a computer program product tangibly embodied in a computer-readable storage medium.

Abstract: The execution of a process is analyzed by a data processing method using a data analysis device for processing a job by combining the processes of transforming input data and generating output data. The execution history of the process is recorded, and each time the process is executed, it is determined whether the process execution has a high anomaly probability based on a predetermined determination indicator. Upon determining that the process execution has a high anomaly probability, the process log information of the process including the determination indicator is recorded, and based on the process log information and the process execution history, the process with the high anomaly probability is further analyzed. By determining whether the process is anomalous or not, the process history analysis information including the analysis result and the execution history of the process to be analyzed is generated.

Abstract: Systems and methods for automatically processing a textual document by identifying occurrences of a piece of text having a predetermined format in the textual document; determining a depth-of-treatment value for each piece of formatted text in the textual document, the depth-of-treatment value indicating a depth of treatment in the textual document afforded to the particular piece of formatted text; associating an abstract with each piece of formatted text in the textual document; and generating a data record containing each identified piece of formatted text from the textual document, the depth-of-treatment value and the abstract associated with each piece of formatted text in the textual document.

Abstract: Systems and methods for finding related terms based on three different sources are disclosed. Generally, a first plurality of distances is determined based on one or more received terms and a first plurality of terms derived from an algorithmic search list. A second plurality of distances is determined based on the one or more received terms and a second plurality of terms derived from a sponsored search list. A third plurality of distances is determined based on the one or more received terms and a third plurality of terms derived from search logs. The first, second, and third pluralities of distances are combined to derive a fourth plurality of distances. Finally, a plurality of related terms related to the one or more received terms is generated based on the fourth plurality of distances.

Abstract: An improved duplicate detection technique that uses query-relevant information to limit the portion(s) of documents to be compared for similarity is described. Before comparing two documents for similarity, the content of these documents may be condensed based on the query. In one embodiment, query-relevant information or text (also referred to as “snippets”) is extracted from the documents and only the extracted snippets, rather than the entire documents, are compared for purposes of determining similarity.

Abstract: In one embodiment, an original search input entered is detected. A plurality of stored terms in a record are detected. The original search input is modified to remove an excluded term from the search input to form a modified search input. The modified search input is matched against the plurality of stored terms to form a first result that includes a number of the stored terms. The number of the stored terms in the first result is compared to a threshold value. In response to the number of the stored terms in the first result satisfying the threshold value, the stored terms of the first result are displayed. Otherwise, the original search input is matched against the plurality of stored terms to form a second result that includes a number of the stored terms, and the stored terms of the second result are displayed.