Abstract: Systems and methods for enhanced document analysis and identification through a reinforcement learning framework are provided. A system may employ computer based reinforcement learning to interact with large populations of documents to help users achieve a goal. To accomplish this goal, rewards, value functions, states, policies, and actions may be modeled and various tools within the system can be used to achieve the user's goal. As actions are performed, the results of these actions may be used to update state, assess rewards, and update value functions and policy functions. If the goal is not achieved, the system may make adjustments by adjusting policies, pushing the user closer to their goal by methods of reinforcement learning. Once the goal is achieved, such as a confidence that at least a certain percentage of relevant documents have been identified, relevant documents may be provided to a party desiring the documents.

Abstract: Described is a new framework that combines the best of both exploratory search and social search: Collaborative exploratory search. This system allows a small group of focused information seekers to search through a collection of information in concert. The system provides exploratory feedback not only based on the individual's search behavior, but on the current, active search behavior of one's fellow search allies. The assumption is that the users who have gotten together to search collaboratively have the same information need, but differing perspectives and insights as to how to best express the queries to meet that need. The collaborative exploratory search system will therefore provide tools and algorithmic support to focus, enhance, and augment searcher activities. Searchers can, by interacting with each other through system-mediated information displays, help each other find all relevant information more efficiently and effectively.

Abstract: A search system generates a plurality of pseudo-documents for a set of documents. Each pseudo-document includes a representation one or more documents from the set of documents that would be retrieved in response to executing a respective basis query over the set of documents. The search system generates an index of the plurality of pseudo-documents. The index provides an indication, for a plurality of respective documents from the set of documents, of the pseudo-documents that include a representation of the respective document. After generating the index, the search system receives a document-query that includes a representation of a first document; and in response to receiving the document-query, the search system identifies one or more pseudo-documents that include representations of at least one document having a predefined relationship to the first document; and generates a result that is based on the first set of one or more pseudo-documents.

Abstract: Speculative query expansion is a way to increase the effectiveness of relevance feedback. Instead of picking the best expansion query beforehand, multiple queries are executed and the best expansion query is selected afterward, but with no additional effort (or information) required from the user. From the best expansion query, search results and expanded queries more pertinent to the user can therefore be provided.

Abstract: Systems and methods are directed to use-based product searching. Raw product information data, such as product features, specifications, and user reviews are processed and analyzed using pattern-based text analysis to extract relevant product aspects and uses. The aspects are weighted in relation to their importance for various uses, and the corresponding aspects and their weights are linked to the uses. A user selects uses for a product, which correspond to weighted aspects, and the weights for the aspects are used to rank products using the weights of the aspects linked to the selected uses. The ranked products are presented to the user in a customizable interface. The user may directly specify weights for the extracted aspects to further customize the ranked list of products. The interface provides additional options for viewing product details, opinions and comparisons.

Abstract: Speculative query expansion is a way to increase the effectiveness of relevance feedback. Instead of picking the best expansion query beforehand, multiple queries are executed and the best expansion query is selected afterward, but with no additional effort (or information) required from the user. From the best expansion query, search results and expanded queries more pertinent to the user can therefore be provided.

Abstract: A search system generates a plurality of pseudo-documents for a set of documents. Each pseudo-document includes a representation one or more documents from the set of documents that would be retrieved in response to executing a respective basis query over the set of documents. The search system generates an index of the plurality of pseudo-documents. The index provides an indication, for a plurality of respective documents from the set of documents, of the pseudo-documents that include a representation of the respective document. After generating the index, the search system receives a document-query that includes a representation of a first document; and in response to receiving the document-query, the search system identifies one or more pseudo-documents that include representations of at least one document having a predefined relationship to the first document; and generates a result that is based on the first set of one or more pseudo-documents.

Abstract: A novel method for going beyond the observed properties of a keyword, to a model in which the presence of a term in a document is assessed not by looking at the actual occurrence of that term, but by a set of non-independent supporting terms, defining the context. In other words, similarity is determined not by properties of the keyword, but by properties of the keyword's context. This yields a scoring for documents which is useful for ad hoc retrieval and, by extension, any information retrieval task where keyword-based similarity is needed.

Abstract: A method and system for additively, manually sequentially or manually in parallel, selecting multiple points or regions of interest in the analog or digital viewfinder of a photo or video camera with automatic controls like auto-focusing, auto-metering and auto white-balancing. Points or regions capable of being selected, within a grid of elements comprising a camera's evaluative matrix and represented to the user in a viewfinder or other display, may be hard-coded (pre-defined, selectable regions) or any point, pixel, or group of pixels. Data from the selected regions is used to simultaneously, dynamically adjust focus related parameters, light metering and white balancing of the camera and is tied into these features. Point selection may be via touch screen or by holding down a lock button. The selected points or regions may be applied to focus and aperture related adjustments or for any other camera function related to measurement of the scene.

Abstract: Described is a component-based control system involving the interface and middleware layers for collaborative exploratory search. The components include modules for multi-user input and display capabilities, and are individually configurable to allow simultaneous manipulation of multiple search parameters and algorithms. In a collaborative exploratory search, a team of people with a shared information need engage in exploratory search together. This search happens synchronously, leveraging realtime feedback in the search loop. The search team works together, finding patterns and information that each player individually might not have found, and doing so more efficiently than any single person could have. Each team member brings their own expertise and point of view to a shared problem. Distributing tasks and roles among team members leverages individual expertise and creates efficiencies of scale.

Abstract: A method and system for additively, manually sequentially or manually in parallel, selecting multiple points or regions of interest in the analog or digital viewfinder of a photo or video camera with automatic controls like auto-focusing, auto-metering and auto white-balancing. Points or regions capable of being selected, within a grid of elements comprising a camera's evaluative matrix and represented to the user in a viewfinder or other display, may be hard-coded (pre-defined, selectable regions) or any point, pixel, or group of pixels. Data from the selected regions is used to simultaneously, dynamically adjust focus related parameters, light metering and white balancing of the camera and is tied into these features. Point selection may be via touch screen or by holding down a lock button. The selected points or regions may be applied to focus and aperture related adjustments or for any other camera function related to measurement of the scene.

Abstract: Personalized change-tracking functionality for collaboratively authored documents is provided. Initially, every document or page for a given user is marked as “new”, “unread” or “unaccessed”. Any time a user edits or even reads a portion of a document, that fact is captured and those segments of the document are marked, for only that user, as “read”. However, any time another user edits that same document, if those edits overwrite areas that the first user has read or edited, or open up new areas, all those areas revert to the “new” or “unread” status. Thus, the next time the user visits that document or page, the display module will show him not what is new, but what is new to him.

Abstract: Described is a component-based control system involving the interface and middleware layers for collaborative exploratory search. The components include modules for multi-user input and display capabilities, and are individually configurable to allow simultaneous manipulation of multiple search parameters and algorithms. In a collaborative exploratory search, a team of people with a shared information need engage in exploratory search together. This search happens synchronously, leveraging realtime feedback in the search loop. The search team works together, finding patterns and information that each player individually might not have found, and doing so more efficiently than any single person could have. Each team member brings their own expertise and point of view to a shared problem. Distributing tasks and roles among team members leverages individual expertise and creates efficiencies of scale.

Abstract: A real-time, content-based document navigation and caching tool for use within linked-document environments. The system includes a combination of rooted spidering, ranking and displaying to provide the user with a capability to perform intelligent navigation of the document collection. The system allows users to intelligently browse their local information environments, finding the documents that are most relevant to their information needs but also respecting the locality of the user's current location within the hyperlinked environment.

Abstract: Embodiments of the present invention enable an approach to index segments of a media stream containing of visual and textual information, using a combination of visual, textual, auditory and temporal features to combine segments that correspond to topical contexts into logical groups. A visual/temporal/auditory/textual weighting scheme is adopted, which allows segments from elsewhere in the same media stream to affect the index terms associated with the current segment. This description is not intended to be a complete description of, or limit the scope of, the invention. Other features, aspects, and objects of the invention can be obtained from a review of the specification, the figures, and the claims.

Abstract: A novel method for going beyond the observed properties of a keyword, to a model in which the presence of a term in a document is assessed not by looking at the actual occurrence of that term, but by a set of non-independent supporting terms, defining the context. In other words, similarity is determined not by properties of the keyword, but by properties of the keyword's context. This yields a scoring for documents which is useful for ad hoc retrieval and, by extension, any information retrieval task where keyword-based similarity is needed.