Microsoft Technology Licensing, LLC

Other

Eric David Brill

Eric J Horvitz

Susan Theresa Dumais

Robert L Rounthwaite

Gary William Flake

Joshua T Goodman

Oliver Hurst-Hiller

Ryen William White

Alexander G Gounares

Christopher A Meek

Raymond Edward Ozzie

Robert J Ragno

Matthew R Richardson

Robert C Moore

Christopher J C Burges

Ramez Naam

Silviu-Petru Cucerzan

Trenholme J Griffin

David R Azari

Bradly A Brunell

Allison P Heath

Surajit J Chaudhuri

Anoop Gupta

John Carlton Platt

Stephen L Lawler

Arul A Menezes

Raman Chandrasekar

Christopher J Burges

William Henry Gates, Iii

Xuedong D Huang

Steven Mark Drucker

Josh D Benaloh

Arnd Christian König

Kyle G Peltonen

Jesper B Lind

Hugo Zaragoza

Marc A Smith

Jamie Paul Buckley

Simon H Corston

Christopher D Payne

Randall Kern

Luis Von Ahn

Duncan L Davenport

Yevgeny E Agichtein

Wensi Xi

Silviu Petru Cucerzan

Laura J Baldwin

Harr Chen

Noah A Smith

Serguei Vassilvitskii

Philip F Carmichael

Harold C Daume, Iii

Soumya Ray

Aspects of the subject matter described herein relate to predicting and using search engine switching behavior. In aspects, switching components receive a representation of user interactions with at least one browser. The switching components derive information from the representation that is useful in predicting whether a user will switch search engines. The derived information and information about a user's current interaction with a browser is then used by a switch predictor to predict whether the user will switch search engines. This prediction may be used in a variety of ways examples of which are given herein.

Technologies for encouraging search engine switching based on behavior patterns

A question processing system routes questions among participants in a balanced and sustainable manner. Any participant can act as an inquirer (who poses questions) and an expert (who answers questions). In one illustrative case, the question processing system operates by: receiving a question from an inquirer; determining at least one expert that is appropriate to answer the question; and routing the question to the expert. The receiving, determining, and routing are repeated with respect to other inquirers and other experts to achieve a desired balance of information exchange among the plurality of participants of the electronic question processing system.

Balanced routing of questions to experts

Architecture for generating responses to search engine queries. Question/answer pairs are used to provide a predefined question hierarchy to facilitate a search. Informative terms and one or more questions are associated with each question node of the hierarchy. A question posed is mapped into the hierarchy to the most similar question class node, as determined by passages of the question and/or comparing the informative terms of the question and the question/answer node. Weighted informative terms associated with that node are then returned, and can be used by an automatic question answering system to more effectively find an answer to a question.

Architecture for generating responses to search engine queries

Architecture is provided that facilitates user-controlled access to user profile information related to spam filtering. A user is allowed to selectively expose (or mask) portions of his/her profile related to spam filtering to third parties. Additionally, advertisers and/or content providers can offer incentives or enticement in response to the acceptance of which a user exposes larger portions of their profiles related to spam filtering. The architecture comprises a system that facilitates profile management utilizing a profile component that facilitates creation and storage of an electronic profile including spam filtering information of a user, and a control component under control of the user for controlling access to the profile.

Content presentation based on user preferences

The present invention provides systems and methods that employ statistical filtering and ranking techniques to improve content search engine search results by tuning a general-purpose search engine for an entry point for a group of users. The filter can be manually and/or automatically configured via providing training sets of relevant and non-relevant data. For example, a relevant set of data comprising web pages associated with a desired search context for the group utilizing the entry point can be employed. In addition, a non-relevant set of data comprising random and unrelated documents can be employed. The learned filter can then be employed to compare a returned result with the known relevant and the known non-relevant data sets to determine whether the result is relevant to the user for the entry point, and the degree of relevance. The results can then be presented to the user based on the relevance.

Systems and methods to tune a general-purpose search engine for a search entry point

Predicting and using search engine switching behavior

Architecture is provided that facilitates user-controlled access to user profile information. A user is allowed to selectively expose (or mask) portions of his/her profile to third parties. Additionally, advertisers and/or content providers can offer incentives or enticement in response to the acceptance of which a user exposes larger portions of their profile. The architecture comprises a system that facilitates profile management utilizing a profile component that facilitates creation and storage of an electronic profile of a user, and a control component under control of the user for controlling access to the profile. Machine learning and reasoning is provided to make inferences and automate aspects thereof.

A context based search and document retrieval application is disclosed. The application uses context data that is generated by a document analysis component based upon context information from various sources, including for example, document collections available on the world wide web. The context data is provided to a search query generator. The search query generator is configured to generate a contextually annotated search query based upon a user query and context data. Bookmarks having an associated context are generated for documents retrieved in response to the contextually annotated search query. The application then provides access to the bookmarked documents when the user is browsing other documents with similar contexts to the context of the original document for which the context based search was executed and the bookmark created. The bookmarked documents are hyperlinked to terms in the browsed document.

Context based search and document retrieval

A system that can analyze a multi-dimensional input thereafter establishing a search query based upon extracted features from the input. In a particular example, an image can be used as an input to a search mechanism. Pattern recognition and image analysis can be applied to the image thereafter establishing a search query that corresponds to features extracted from the image input. The system can also facilitate indexing multi-dimensional searchable items thereby making them available to be retrieved as results to a search query. More particularly, the system can employ text analysis, pattern and/or speech recognition mechanisms to extract features from searchable items. These extracted features can be employed to index the searchable items.

Visual and multi-dimensional search

The claimed subject matter provides systems and methods that facilitate providing specialized content associated with web results in response to a user query to provide queried information directly to a user without requiring the user to select a link and/or scroll through information on a web page associated therewith. Reference language models can be generated from multiple queries and can be compared to a query language model generated from web results returned in response to a user query. Based on such comparison, a determination can be made regarding whether a particular portion of specialized content corresponding to some particular query keywords is relevant enough to warrant direct presentation to the user.

Arbitration of specialized content using search results

A unique multi-stage classification system and method that facilitates reducing human resources or costs associated with text classification while still obtaining a desired level of accuracy is provided. The multi-stage classification system and method involve a pattern-based classifier and a machine learning classifier. The pattern-based classifier is trained on discriminative patterns as identified by humans rather than machines which allow a smaller training set to be employed. Given humans' superior abilities to reason over text, discriminative patterns can be more accurately and more readily identified by them. Unlabeled items can be initially processed by the pattern-based classifier and if no pattern match exists, then the unlabeled data can be processed by the machine learning classifier. By employing the classifiers in this manner, less human involvement is required in the classification process. Even more, classification accuracy is maintained and/or improved.

Reducing human overhead in text categorization

Structured content and associated metadata from the Web are leveraged to provide specific answer string responses to user questions. The structured content can also be indexed at crawl-time to facilitate searching of the content at search-time. Ranking techniques can also be employed to facilitate in providing an optimum answer string and/or a top K list of answer strings for a query. Ranking can be based on trainable algorithms that utilize feature vectors for candidate answer strings. In one instance, at crawl-time, structured content is indexed and automatically associated with metadata relating to the structured content and the source web page. At search-time, candidate indexed structured content is then utilized to extract an appropriate answer string in response to a user query.

Question answering over structured content on the web

A system that employs an explicitly and/or implicitly trained model in order to return entity-specific computer-based search results is provided. The innovation can provide for a customized search model that focuses search in connection with achieving information that is meaningful with respect to goals of an entity. The model can be used to modify a search query in accordance with a goal of the entity or to generate the search query thereby returning meaningful and/or targeted results to the user. The system can automatically gather entity-related data thereafter determining or inferring a goal as well as training the model. Moreover, the system can selectively configure (e.g., order, rank, filter) and render results to a user based upon the model.

Entity-specific search model

The subject invention relates to a system and/or methodology that facilitate converting structured data (e.g., tabular information) into natural language, thus making it available to question answering systems and search engines. More specifically, the system and method employ the vast quantities of natural language on the particular storage system, database, network, or server that a user wishes to search. For example, the system can utilize natural-language based information located on the Web to assist in the task of converting structured data into natural language. In particular, the system can find at least one sentence or sentence fragment on the Web that refers to at least one tuple (e.g., row) in a given table. Following, those sentences can be generalized across all or substantially all tuples in the table. The resulting body of text can be used by the question answering system to answer user queries.

Table querying

Growing community of inventors

Redmond, WA, United States of America

Eric David Brill