Yahoo! Inc.

Utku Irmak

Reiner Kraft

Vadim Von Brzeski

The subject matter disclosed herein relates to ranking and/or indexing of documents based at least in part on user interaction with contextual shortcuts. In one or more implementations, data may be collected regarding user interaction with one or more contextual shortcuts and a query may be received from a user based at least in part on an activation of a given contextual shortcut. A search result may be transmitted to the user via a contextual search interface based at least in part on the collected data regarding user interaction with one or more contextual shortcuts.

Ranking documents through contextual shortcuts

Disclosed are methods and apparatus for performing named entity recognition. A set of candidates and corresponding contexts are obtained, each of the set of candidates being a potential seed example of an entity. The contexts of at least a portion of the set of candidates are compared with contexts of a set of seed examples of the entity such that a subset of the set of candidates are added to the set of the seed examples. A set of rules are created from the set of seed examples obtained in the comparing step. A final set of seed examples of the entity is generated by executing the set of rules against the set of candidates.

Scalable semi-structured named entity detection

Techniques are provided for ranking the entities that are identified in a document based on an estimated likelihood that a user will actually make use of the annotations. According to one disclosed approach, usage data that indicates how users interact with annotations contained in documents presented to the users is collected. Based on the usage data, weights are generated for features of a feature vector. The weights are then used to modify feature scores of entities, and the modified feature scores are used to determine how to annotate documents. Specifically, a set of entities are identified within a document. A ranking for the identified entities is determined based, at least in part, on (a) feature vector scores for each of the identified entities, and (b) the weights generated for the features of the feature vector. The document is then annotated based, at least in part, on the ranking.

Growing community of inventors

Santa Clara, CA, United States of America

Utku Irmak