Abstract: A service is disclosed for enabling web sites and other entities to provide item recommendations and other behavior-based content to end users. The service can be implemented as a web service that is remotely accessible over the Internet. Web sites use the web service's interface to report events descriptive of item-related actions performed by end users (e.g., item views, item purchases, searches for items, etc.). The web service analyzes the reported event data on an aggregated basis to detect various types of associations between particular items, and stores resulting datasets that map items to associated items. The web service's interface also provides various API calls for enabling the web sites to request item recommendations and other behavior-based content, including but not limited to personalized recommendations that are based on the event history of the target user. Advantageously, the web sites need not host the infrastructure for providing such content.

Abstract: Various methods are disclosed for monitoring user browsing activities that indicate user interests in particular products, or other items, represented in an electronic catalog, and for using such information to identify items that are related to one another. In one embodiment, relationships between items within an electronic catalog are determined by identifying items that are frequently viewed by users within the same browsing session (e.g., items A and B are related because a significant portion of those who viewed A also viewed B). The resulting item relatedness data may be stored in a table that maps items to sets of related items. The table may be used to provide personalized item recommendations to users, and/or to supplement item detail pages of the electronic catalog with lists of related items. In one embodiment, the table is used to provide session-specific item recommendations to users.

Abstract: Data mining systems and methods are disclosed for generating data that is helpful to users in selecting between items represented in an electronic data repository, such as an electronic catalog. One disclosed data mining method generates pairwise comparison data for particular pairs of items. The pairwise comparison data for a given item pair reveals the degree to which users who consider both items in the pair select one item over the other. The pairwise comparison data may be appropriately exposed to users of the electronic repository. For instance, in the context of an electronic catalog, an item detail page for item A may be supplemented with an indication that users who view both item A and item B select item B a specified percentage of the time. Another data mining method uses item viewing histories and item purchase histories of users in combination to identify pairs of items that are good candidates for being recommended in combination.

Abstract: A service is disclosed for enabling web sites and other entities to provide item recommendations and other behavior-based content to end users. The service can be implemented as a web service that is remotely accessible over the Internet. Web sites use the web service's interface to report events descriptive of item-related actions performed by end users (e.g., item views, item purchases, searches for items, etc.). The web service analyzes the reported event data on an aggregated basis to detect various types of associations between particular items, and stores resulting datasets that map items to associated items. The web service's interface also provides various API calls for enabling the web sites to request item recommendations and other behavior-based content, including but not limited to personalized recommendations that are based on the event history of the target user.

Abstract: Strategies are described for generating recommendations. The strategies generate a set of original recommendations based on a source of information. The strategies then transform the set of original recommendations into a set of similarity-spaced recommendations based on “repulsion force” analysis applied to the set of original recommendations. In a first implementation, the set of spaced recommendations represent a diverse sampling of items in the set of original recommendations. In a second implementation, the set of spaced recommendations represent a sampling of items in the set of original recommendations which omits or excludes recommendations assessed as obvious. A third implementation can combine the first and second implementations.

Abstract: Strategies are described for generating recommendations. The strategies generate a set of original recommendations based on a source of information. The strategies then transform the set of original recommendations into a set of similarity-spaced recommendations based on “repulsion force” analysis applied to the set of original recommendations. In a first implementation, the set of spaced recommendations represent a diverse sampling of items in the set of original recommendations. In a second implementation, the set of spaced recommendations represent a sampling of items in the set of original recommendations which omits or excludes recommendations assessed as obvious. A third implementation can combine the first and second implementations.

Abstract: A system provides recommendations of web sites, web pages, and/or products to a user based on web pages viewed during a current browsing session. In one embodiment, a browser plug-in or other client program monitors and reports information regarding browsing activities of users across multiple web sites. The resulting cross-site browse histories of the users are analyzed on an aggregated basis to detect behavior-based associations between particular sites, pages and/or products. The detected associations are in turn used to provide personalized recommendations to users. The associations and recommendations may also be based on an automated analysis of the content of the web pages represented in the users' browse histories.

Abstract: A service is disclosed for enabling web sites and other entities to provide item recommendations and other behavior-based content to end users. The service can be implemented as a web service that is remotely accessible over the Internet. Web sites use the web service's interface to report events descriptive of item-related actions performed by end users (e.g., item views, item purchases, searches for items, etc.). The web service analyzes the reported event data on an aggregated basis to detect various types of associations between particular items, and stores resulting datasets that map items to associated items. The web service's interface also provides various API calls for enabling the web sites to request item recommendations and other behavior-based content, including but not limited to personalized recommendations that are based on the event history of the target user. Advantageously, the web sites need not host the infrastructure for providing such content.

Abstract: A service is disclosed for enabling web sites and other entities to provide item recommendations and other behavior-based content to end users. The service can be implemented as a web service that is remotely accessible over the Internet. Web sites use the web service's interface to report events descriptive of item-related actions performed by end users (e.g., item views, item purchases, searches for items, etc.). The web service analyzes the reported event data on an aggregated basis to detect various types of associations between particular items, and stores resulting datasets that map items to associated items. The web service's interface also provides various API calls for enabling the web sites to request item recommendations and other behavior-based content, including but not limited to personalized recommendations that are based on the event history of the target user. Advantageously, the web sites need not host the infrastructure for providing such content.

Abstract: Methods of organizing a series of sibling data entities in a digital computer are provided for preserving sibling ranking information associated with the sibling data entities and for attaching the sibling ranking information to a joint parent of the sibling data entities to facilitate on-demand generation of ranked parent candidates. A rollup function of the present invention builds a rollup matrix (126) that embodies information about the sibling entities and the sibling ranking information and provides a method for reading out the ranked parent candidates from the rollup matrix in order of their parent confidences (141). Parent confidences are based on the sibling ranking information, either alone or in combination with n-gram dictionary ranking or other ranking information.

Abstract: Methods of organizing a series of sibling data entities in a digital computer are provided for preserving sibling ranking information associated with the sibling data entities and for attaching the sibling ranking information to a joint parent of the sibling data entities to facilitate on-demand generation of ranked parent candidates. A rollup function of the present invention builds a rollup matrix (126) that embodies information about the sibling entities and the sibling ranking information and provides a method for reading out the ranked parent candidates from the rollup matrix in order of their parent confidences (141). Parent confidences are based on the sibling ranking information, either alone or in combination with n-gram dictionary ranking or other ranking information.

Abstract: A controlled database system is designed for rapid access of data records with reduced memory storage requirements. The database system employs a set of obfuscated data records stored in data crystals that can only be accessed and read by an iterator, which is not directly accessible by the users of the database. The iterator accesses information responsive to a predefined query sent from a customer application. Rather than providing general tools to customers for constructing any possible queries, such as is done in structured query language database systems, database systems embodying the present invention allow only predefined types of queries to be used by customer applications. By restricting the types of queries customer applications can call, valuable data records remain secure from unauthorized reconstruction or duplication while still allowing limited access for specific purposes.

Abstract: Various methods are disclosed for monitoring user browsing activities that indicate user interests in particular products or other items, and for using such information to identify items that are related to one another. In one embodiment, relationships between products within an online catalog are determined by identifying products that are frequently viewed by users within the same browsing session (e.g., products A and B are related because a significant portion of those who viewed A also viewed B). The resulting item relatedness data is preferably stored in a table that maps items to sets of related items. The table may be used to provide personalized product recommendations to users, and/or to supplement product detail pages with lists of related products. In one embodiment, the table is used to provide session-specific product recommendations to users that are based on the products viewed by the user during the current browsing session.

Abstract: Various methods are disclosed for monitoring user browsing activities, and for using such information to provide session-specific item recommendations to users. In one embodiment, a monitoring component of a merchant's Web site maintains a record of products viewed by each user during a current browsing session—preferably based on visits to product detail pages. A recommendations component uses the resulting history of viewed products to identify additional products to recommend, preferably using a pre-existing table that maps products to related products. In one embodiment, this table is generated by periodically analyzing user browsing histories to identify correlations between purchases, viewing events, and/or other actions performed with respect to particular products. The recommended items may be displayed together with an option to individually deselect the recently viewed items on which the recommendations are based.

Abstract: A controlled database system is designed for rapid access of data records with reduced memory storage requirements. The database system employs a set of obfuscated data records stored in data crystals that can only be accessed and read by an iterator, which is not directly accessible by the users of the database. The iterator accesses information responsive to a predefined query sent from a customer application. Rather than providing general tools to customers for constructing any possible queries, such as is done in structured query language database systems, database systems embodying the present invention allow only predefined types of queries to be used by customer applications. By restricting the types of queries customer applications can call, valuable data records remain secure from unauthorized reconstruction or duplication while still allowing limited access for specific purposes.

Abstract: A limited-access database system is designed for rapid access of data records with reduced memory storage requirements. The database system employs a set of obfuscated data records stored in data crystals that can only be accessed and read by an iterator, which is not directly accessible by the users of the database. The iterator accesses information responsive to a predefined query sent from a customer application. Rather than providing general tools to customers for constructing any possible queries, such as is done in structured query language database systems, database systems embodying the present invention allow only predefined types of queries to be used by customer applications. By restricting the types of queries customer applications can call, valuable data records remain secure from unauthorized reconstruction or duplication while still allowing limited access for specific purposes.

Abstract: Methods of organizing a series of sibling data entities in a digital computer are provided for preserving sibling ranking information associated with the sibling data entities and for attaching the sibling ranking information to a joint parent of the sibling data entities to facilitate on-demand generation of ranked parent candidates. A rollup function of the present invention builds a rollup matrix (126) that embodies information about the sibling entities and the sibling ranking information and provides a method for reading out the ranked parent candidates from the rollup matrix in order of their parent confidences (141). Parent confidences are based on the sibling ranking information, either alone or in combination with n-gram dictionary ranking or other ranking information.

Abstract: Methods of organizing a series of sibling data entities in a digital computer are provided for preserving sibling ranking information associated with the sibling data entities and for attaching the sibling ranking information to a joint parent of the sibling data entities to facilitate on-demand generation of ranked parent candidates. A rollup function of the present invention builds a rollup matrix (126) that embodies information about the sibling entities and the sibling ranking information and provides a method for reading out the ranked parent candidates from the rollup matrix in order of their parent confidences (141). Parent confidences are based on the sibling ranking information, either alone or in combination with n-gram dictionary ranking or other ranking information.

Abstract: This invention relates to the field of optical character recognition for data in drop-out forms. In the absence of permanent, fixed landmarks, this invention uses the structure of the entered data to provide a landmark to be used in solving various problems encountered in automated processing of drop-out forms. An expected data pattern is located in a scanned image and compared to a defined data pattern in a template form that serves as a reference.

Abstract: A system provides session-specific web page or web site recommendations to a user based upon an identification of web pages previously viewed by the user during a browsing session. During a sequence of proximately visited locations, users tend to view web pages with similar content. To collect data, a client program executes in conjunction with a web browser on each of multiple users' computers. Each client program identifies pages viewed by the user and transmits the sequence of identifications to a server application executing on a recommendation system. The recommendation system creates tables of similar web pages based upon the sequences of locations visited by users. To create session-specific web page recommendations, the system uses the client program to identify a set of locations visited by the user during the session. The system then identifies similar web pages based upon the created tables and combines, sorts, and filters the results.