Google Inc.

Diego Federici

Kai Chen

Steven G Dropsho

Charles Johan Larsson Tibell

Nitin Kumar Gupta

Steven D Baker

Srinivasan Venkatachary

Michelangelo Diligenti

Yi Liu

Henrik Jacobsson

Per Bjornsson

Lingkun Chu

Robert Andrew Brennan

In general, in one aspect, a method includes compiling user interaction statistics for a set of content items displayed in association with a first target media document having a non-textual portion, at least some of the content items associated with one or more keywords, based on the interaction statistics, associating the first target media document with at least some of the keywords associated with the content items, and based on a common attribute of the first target media document and a second target media document having a non-textual portion, associating the second target media document with at least some of the keywords assigned to the first target media document. Other aspects include corresponding systems, apparatus, and computer programs stored on computer storage devices.

Content keyword identification

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using search entity transition probabilities. In some implementations, data identifying entities and transition probabilities between entities is stored in a computer readable medium. Each transition probability represents a strength of a relationship between a pair of entities as they are related in search history data. In some implementations, an increase in popularity for a query is identified and a different query is identified as temporally related to the query. Scoring data for documents responsive to the different query is modified to favor newer documents. In other implementations, data identifying a first session as spam is received, and a spam score is calculated for either a second session of queries or a single query using transition probabilities. The second session (or single query) is identified as spam from the spam score.

Search entity transition matrix and applications of the transition matrix

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating candidate answer passages. In one aspect, a method includes receiving a query determined to be a question query data identifying resources determined to be responsive to the query; for each resource in a top-ranked subset of the resources: identifying a plurality of passage units in the resource; applying a set of passage unit selection criterion to the passage units, each passage unit selection criterion specifying a condition for inclusion of a passage unit in a candidate answer passage, wherein a first subset of passage unit selection criteria applies to structured content and a second subset of passage unit selection criteria applies to unstructured content; and generating, from passage units that satisfy the set of passage unit selection criterion, a set of candidate answer passages.

Candidate answer passages

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for determining similarities between queries. A first query and a second query are identified, and first related queries associated with the first query and second related queries associated with the second query are also identified. One or more first features for the first query are identified, and one or more second features for the second query are also identified. A first score is calculated for each first feature based on the relatedness weight associated with the respective first related query including the term of the respective first feature, and a second score is calculated for each second feature based on the relatedness weight associated with the respective second related query including the term of the respective second feature. A similarity of the first query and the second query is determined based on the first score and the second score.

Co-click based similarity score of queries and keywords

A computer-implemented method for processing network activities is described. The method includes identifying a model that specifies attributes for network objects, identifying a network object having one or more attributes that deviate from the model, and providing as an input to a ranking algorithm a value associated with the deviance of the one or more attributes of the identified network object.

Growing community of inventors

Zurich, Switzerland

Diego Federici