Abstract: A conversational recommender system retrieves k cases (3) and generates a structure for user feedback (4). The structure includes a list of critique units for each of which the user can specify a value range. It also generates compound critiques, each being a combination of features and value ranges presented together. An explanation of a compound critique can be generated in response to a user request. The system determines (7), from one cycle of a session to the next, if a preference has been carried. If so, a re-focus function implementing diversity is used for retrieval in the next cycle. If not, a refine function implementary similarity-based retrieval is used for the next cycle of the session.

Abstract: A system has a primary server storing search data sets (“staks”) and a linked social network server. Interaction with the server is via a client and a Web site. There is a software agent code executing on the client. The software agent component provides full integration with underlying search engines so that users can continue to search in the normal way, using their favourite search engines, while benefiting from management of search staks, automatic stak selection, and result promotion. The system allows users to manage their staks and providing a range of social networking style services to help users make contact with other like-minded searchers. In addition, it allows users to search for relevant staks. The software agent component provides browser-based access to allow users to manage and share their searchers direct from their browser, as well as providing the benefits of search promotions as they search normally.

Abstract: A meta search engine receives a query from a user, and submits adapted queries to n search engines. The results are adapted and combined to provide a result set R?. In parallel, the meta search engine accesses a selected hit matrix populated with values for correspondence of queries q1-qm with web pages pi-pn. These values are the number of “hits” for pervious searches. The pages retrieved from a row for qT are combined with the search engine results to provide a final result but which is ranked according to previous search experience. The hit matrix may be dynamically selected according to a user community. The query need not exactly match a hit matrix query as rows can be selected according to similarity to the current query qT with weightings applied.

Abstract: In an information retrieval method a meta search engine (4) receives a user information query (qT) in the form of key words; and submits the query to a search engine (2), and also submits the query to a snippet database (6) storing previously selected results, indexed by a snippet index of terms included in snippets of the results when they were previously selected. The meta search engine (4) then receives results from the search engine (2); and receives a response from the snippet database (6). It merges the results from the search engine (2) with those from the snippet database (6); and submits re-ranked results to the user. The snippet database (5, 6) processes in parallel with the underlying search engine (2). The meta search engine (4) also monitors which results of the returned re-ranked results are selected by the user, and updates the snippet index accordingly.

Abstract: A meta search engine (1) receives a query (qT) from a user (U), and submits adapted queries (q1-qn) to n search engines (S1-Sn). The results (R1-Rn) are adapted (R1?-Rn?) and combined to provide a result set R?. In parallel, the meta search engine (1) accesses a selected hit matrix (H) populated with values for correspondence of queries q1-qm with web pages pi-pn. These values are the number of “hits” for pervious searches. The pages (R?) retrieved from a row for qT are combined with the search engine (S1-Sn) results to provide a final result but which is ranked according to previous search experience. The hit matrix may be dynamically selected according to a user community. The query (qT) need not exactly match a hit matrix query (q1-qn) as rows can be selected according to similarity to the current query qT with weightings applied.

Abstract: A conversational recommender system retrieves k cases (3) and generates a structure for user feedback (4). The structure includes a list of critique units for each of which the user can specify a value range. It also generates compound critiques, each being a combination of features and value ranges presented together. An explanation of a compound critique can be generated in response to a user request. The system determines (7), from one cycle of a session to the next, if a preference has been carried. If so, a re-focus function implementing diversity is used for retrieval in the next cycle. If not, a refine function implementary similarity-based retrieval is used for the next cycle of the session.

Abstract: A method of providing information in a directory system comprising a directory for company information, goods and services; a communications network; a directory computer operated by a directory provider; a plurality of search computers operated by enquirers which search computers are connected within the network; and a plurality of active content providers operating active content provider computers, at least some of which are connected to the directory computer. The company information, goods and services are classified by unique identifiers. Lists of goods and services and company information are provided as provider content from both active and passive content providers. These are stored in a provider database with unique identifiers. The directory computer on request downloads to the search computer one or more queries to allow the search computer specify the search correctly and then the directory computer downloads the search result.

Abstract: A method for personalizing content such as user interface items provided to a user during a session of an interactive application is disclosed. One or more environmental context attribute values for the session are determined. Responsive to a request for content, for at least one item of the content, each of the environmental context attribute values ate mapped to respective values indicating a users level of interest in the item for the environmental context attribute values. Content is retrieved from a content database and the content is personalized as a function of the users level of interest in the item for the environmental context attribute values before being returned to the application.

Abstract: This invention refers to an automatic design process comprising the steps of:- (1)writing a target specification of an item to be designed to a partitioned section of a memory circuit; (2)automatically determining an address for a case stored in a case-base by reference to the target specification;(3)retrieving the case from the case-base and writing the retrieved case to a partitioned section of the memory circuit; and (4)adapting the retrieved case into conformity with the target specification, and writing the adapted case to a partitioned section of the memory circuit as a target solution. Preferably, the process further comprises successive cycles of retrieving a case and adapting the retrieved case, each cycle being initiated by writing of a target specification representing a component of the item to the target specification section of the memory circuit.