Google Inc.

Fernando Pereira

Amarnag Subramanya

John Blitzer

Rahul Gupta

Ciprian Ioan Chelba

Jakob David Uszkoreit

Engin Cinar Sahin

Dekang Lin

Peng Xu

Dipanjan Das

Ni Lao

Hrishikesh Balkrishna Aradhye

Fadi Biadsy

Lev Finkelstein

Douglas Eck

Daniel Martin Bikel

Slav Petrov

Keith Hall

Marius Pasca

Jennifer Ann Gillenwater

Terry Yang-Hoe Koo

Garth Griffin

Rajhans Samdani

Kapil R Thadini

Maria Shugrina

Implementations include systems and methods for querying a data graph. An example method includes receiving a machine learning module trained to produce a model with multiple features for a query, each feature representing a path in a data graph. The method also includes receiving a search query that includes a first search term, mapping the search query to the query, and mapping the first search term to a first entity in the data graph. The method may also include identifying a second entity in the data graph using the first entity and at least one of the multiple weighted features, and providing information relating to the second entity in a response to the search query. Some implementations may also include training the machine learning module by, for example, generating positive and negative training examples from an answer to a query.

Querying a data graph using natural language queries

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for providing app store search results. An example method includes responsive to a first search query directed to an app store: revising the first search query to produce a second search query different from the first search query; obtaining, from an Internet search engine, second search results responsive to the second search query; analyzing the second search results to identify apps available on the app store that are relevant to the second search query; obtaining, from the app store, first search results responsive to the first search query that identify apps available in the app store; and modifying the first search results based on analyzing the second search results.

Providing app store search results

A language processing system collects similar queries and respective responses and aggregated by responses. Incorrect responses are determined and filtered by the aggregation. The remaining responses are then used to query a high precision system for attributes of entities specified by the queries. The attribute type is determined from the responses of the high precision system, and corresponding parse rules are generated. The parse rules are then associated with an operation that yields a response that specifies an attribute of the attribute type.

Inducing command inputs from high precision and high recall data

Systems and techniques are provided for training a natural language processing model with information retrieval model annotations. A natural language processing model may be trained, through machine learning, using training examples that include part-of-speech tagging and annotations added by an information retrieval model. The natural language processing model may generate part-of-speech, parse-tree, beginning, inside, and outside label, mention chunking, and named-entity recognition predictions with confidence scores for text in the training examples. The information retrieval model annotations and part-of-speech tagging in the training example may be used to determine the accuracy of the predictions, and the natural language processing model may be adjusted. After training, the natural language processing model may be used to make predictions for novel input, such as search queries and potential search results. The search queries and potential search results may have information retrieval model annotations.

Training a natural language processing model with information retrieval model annotations

A method identifies pairs of first and second command inputs from respective user device sessions for which the first and second operation data are indicative of a first operation failure and a second operation success. The first operation data indicate a first operation performed on data from a first resource property in response to the first command input, and the second operation data indicate a second operation performed on data from a second resource property in response to the second command input. They system determines, from the identified pairs of first and second command inputs, command inputs for which a parsing rule that is associated with the second operation is to be generated.

Inducing command inputs from property sequences

Speech recognition techniques may include: receiving audio; identifying one or more topics associated with audio; identifying language models in a topic space that correspond to the one or more topics, where the language models are identified based on proximity of a representation of the audio to representations of other audio in the topic space; using the language models to generate recognition candidates for the audio, where the recognition candidates have scores associated therewith that are indicative of a likelihood of a recognition candidate matching the audio; and selecting a recognition candidate for the audio based on the scores.

Speech recognition using topic-specific language models

Implementations include systems and methods generate data for training or evaluating semantic analysis engines. For example, a method may include receiving documents from a corpus that includes an authoritative set of documents from an authoritative source. Each document in the authoritative set may be associated with an entity. A second set of documents from the corpus that do not overlap with the first set may include at least one link to a document in the authoritative set, the at least one link being associated with anchor text. For each document in the second set, the method may include identifying entity mentions in the document based on the anchor text. The method may include associating the entity mention with the entity in a graph-structured knowledge base or associating entity types with the entity mention. The method may also include training a semantic analysis engine using the identified entity mentions and associations.

Automatic annotation for training and evaluation of semantic analysis engines

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training acoustic models. Speech data and data identifying a transcription for the speech data are received. A phonetic representation for the transcription is accessed. Training sequences are identified for a particular phone in the phonetic representation. Each of the training sequences includes a different set of contextual phones surrounding the particular phone. A partitioning key is identified based on a sequence of phones that occurs in each of the training sequences. A processing module to which the identified partitioning key is assigned is selected. Data identifying the training sequences and a portion of the speech data are transmitted to the selected processing module.

Training acoustic models using distributed computing techniques

The disclosure includes a system and method for generating audio snippets from a subset of audio tracks. In some embodiments an audio snippet is an audio summary of a group or collection of songs.

System and method for audio snippet generation from a subset of music tracks

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for graph-based semi-supervised learning of structured tagging models. In one aspect, a method includes creating a graph having a plurality of unique vertices in which vertices in a first set of vertices represent n-grams that are each associated with a respective part-of-speech and that were derived from labeled source domain text, and in which vertices in a different second set of vertices represent n-grams that are not associated with a part-of-speech and that were derived from unlabeled target domain text. A respective measure of similarity is calculated between the vertices in each of the pairs based at least partially on a distance between the respective features of the pair used to weight a graph edge between the pair.

Graph-based semi-supervised learning of structured tagging models

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for recognizing speech using a variable length of context. Speech data and data identifying a candidate transcription for the speech data are received. A phonetic representation for the candidate transcription is accessed. Multiple test sequences are extracted for a particular phone in the phonetic representation. Each of the multiple test sequences includes a different set of contextual phones surrounding the particular phone. Data indicating that an acoustic model includes data corresponding to one or more of the multiple test sequences is received. From among the one or more test sequences, the test sequence that includes the highest number of contextual phones is selected. A score for the candidate transcription is generated based on the data from the acoustic model that corresponds to the selected test sequence.

Speech recognition using variable-length context

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating query suggestions. In one aspect, a method includes receiving candidate text queries and class instance pairs and generating one or more query specializations from the candidate text queries and the class instance pairs. In another aspect, a method includes receiving a partial query entered by a user, identifying one or more query specializations for query completions for the partial query, and presenting query suggestions to the user, including presenting the identified query specializations.

Growing community of inventors

Palo Alto, CA, United States of America

Fernando Pereira