Google Inc.

Microsoft Technology Licensing, LLC

Microsoft Corporation

Ciprian Ioan Chelba

Alejandro Acero

Shumin Zhai

Milind V Mahajan

Yu Ouyang

Fernando Pereira

Johan Schalkwyk

Jakob David Uszkoreit

Maryam Garrett

Peng Xu

Qiang Wu

David Mowatt

Robert L Chambers

Michael Dennis Riley

Xiaojun Bi

Thorsten Brants

John Blitzer

Engin Cinar Sahin

Satoshi Kataoka

Cyril Georges Luc Allauzen

Dekang Lin

Keisuke Kuroyanagi

Rahul Gupta

Asela J R Gunawardana

Ken Wakasa

Yonggang Deng

Edward Oscar Benson

Jorge F Silva Sanchez

Jeffrey Adgate Dean

Sanjay Ghemawat

Brian Strope

Ye-Yi Wang

Hasim Sak

Asela J Gunawardana

Percy Shuo Liang

Daniel Martin Bikel

Brendan John Frey

Erik Warren Selberg

Leon Wong

Roger Peng Yu

Patrick Nguyen

Boulos Harb

Noah B Coccaro

Rachel Imogen Morton

Shiri Azenkot

Preethi Jyothi

Leif Johnson

Frank Torsten B Seide

Yik-Cheung Tam

A computing device outputs for display at a presence-sensitive display, a graphical keyboard comprising a plurality of keys, receives an indication of at least one gesture to select a group of keys of the plurality of keys, and determines at least one characteristic associated with the at least one gesture to select the group of keys of the plurality of keys. The computing device modifies a spatial model based at least in part on the at least one characteristic and determines a candidate word based at least in part on data provided by the spatial model and a language model, wherein the spatial model provides data based at least in part on the indication of the at least one gesture and wherein the language model provides data based at least in part on a lexicon. The computing device outputs for display at the presence-sensitive display, the candidate word.

Touch keyboard using a trained model

Touch keyboard using language and spatial models

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for language modeling of complete language sequences. Training data indicating language sequences is accessed, and counts for a number of times each language sequence occurs in the training data are determined. A proper subset of the language sequences is selected, and a first component of a language model is trained. The first component includes first probability data for assigning scores to the selected language sequences. A second component of the language model is trained based on the training data, where the second component includes second probability data for assigning scores to language sequences that are not included in the selected language sequences. Adjustment data that normalizes the second probability data with respect to the first probability data is generated, and the first component, the second component, and the adjustment data are stored.

Language modeling of complete language sequences

In one example, a method includes: receiving from a first user interface a first input from a first user specifying a first particular instant in a video other than a beginning of the video; in response to the first input, generating by one or more computer systems first data for inclusion in a link to the video, the first data representing the first particular instant in the video and being operable automatically to direct playback of the video at a second user interface to start at the first particular instant in the video in response to a second user selecting the link at the second user interface; and communicating the first data to a link generator for inclusion in the link to the video.

Time-marked hyperlinking to video content

In one example, a method includes outputting, at a presence-sensitive display operatively coupled to a computing device, a graphical keyboard comprising a plurality of keys and receiving an indication of a gesture to select a group of keys of the plurality of keys. The method may further include determining, in response to the indication, a candidate word based at least in part on the group of keys. The method may further include determining a group of alignment points traversed by the gesture and respective cost values for each of at least two keys of the plurality of keys. The method may further include comparing the respective cost values for at least each of at least two keys of the plurality of keys to determine a combination of keys having a combined cost value and outputting the candidate word that is based on the respective cost values.

Incremental feature-based gesture-keyboard decoding

A system is configured to receive a first string corresponding to an interpretation of a natural-language user voice entry; provide a representation of the first string as feedback to the natural-language user voice entry; receive, based on the feedback, a second string corresponding to a natural-language corrective user entry, where the natural-language corrective user entry may correspond to a correction to the natural-language user voice entry; parse the second string into one or more tokens; determine at least one corrective instruction from the one or more tokens of the second string; generate, from at least a portion of each of the first and second strings and based on the at least one corrective instruction, candidate corrected user entries; select a corrected user entry from the candidate corrected user entries; and output the selected, corrected user entry.

Natural language correction for speech input

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

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for using non-parametric models in speech recognition. In some implementations, speech data is accessed. The speech data represents utterances of a particular phonetic unit occurring in a particular phonetic context, and the speech data includes values for multiple dimensions. Boundaries are determined for a set of quantiles for each of the multiple dimensions. Models for the distribution of values within the quantiles are generated. A multidimensional probability function is generated. Data indicating the boundaries of the quantiles, the models for the distribution of values in the quantiles, and the multidimensional probability function are stored.

Speech recognition using non-parametric models

A language processing system identifies sequential command inputs in user session data stored in logs. Each sequence command input is a first command input followed by a second command input. The system determines user actions in response to each command input. For the second command input, an action was taken at the user device in response to the command input, and there is no parsing rule associated with the action that parses to the first command input. If there is a sufficient co-occurrence of the first and second command inputs and the resulting action in the logs, then a parsing rule for the action may be augmented with a rule for the first command input.

Parsing rule augmentation based on query sequence and action co-occurrence

In one example, a method includes outputting, by a computing device and for display, a graphical keyboard comprising a plurality of keys, and receiving, by the computing device, an indication of a gesture detected at a presence-sensitive input device. The method further includes determining, by the computing device, an input posture of the gesture at the presence-sensitive input device, and applying, by the computing device and based at least in part on the input posture, at least one offset to a location associated with a key from the plurality of keys to determine an offset location associated with the key, and in response to receiving the indication of the gesture, selecting, by the computing device and based at least in part on the offset location, the key as a selected key.

Key selection of a graphical keyboard based on user input posture

A data processing apparatus is configured to receive a first string related to a natural-language voice user entry and a second string including at least one natural-language refinement to the user entry; parse the first string into a first set of one or more tokens and the second string into a second set of one or more tokens; determine at least one refining instruction from the second set of one or more tokens; generate, from at least a portion of each of the first string and the second string and based on the at least one refining instruction, a group of candidate refined user entries; select a refined user entry from the group of candidate refined user entries; and output the selected, refined user entry.

Natural language refinement of voice and text entry

A computing device determines, based at least in part on indications of user input, scores for a first set of candidate strings and a second set of candidate strings. Each candidate string from the first set of candidate strings is in a lexicon. Candidate strings from the second set of candidate strings are not necessarily in the lexicon. The computing device determines the scores for the first set of candidate strings based on probabilities of the candidate strings being entered. For each candidate string from the second set of candidate strings, the computing device determines the scores for the candidate string based on probabilities of characters of the candidate string being entered. The computing device selects a candidate string based on the scores for the first and second sets of candidate strings and outputs, for display at the display device, the selected candidate string.

Character and word level language models for out-of-vocabulary text input

Growing community of inventors

Palo Alto, CA, United States of America

Ciprian Ioan Chelba