Knowledge Source Personalization To Improve Language Models

摘要

Systems and methods are provided for improving language models for speech recognition by personalizing knowledge sources utilized by the language models to specific users or user-population characteristics. A knowledge source, such as a knowledge graph, is personalized for a particular user by mapping entities or user actions from usage history for the user, such as query logs, to the knowledge source. The personalized knowledge source may be used to build a personal language model by training a language model with queries corresponding to entities or entity pairs that appear in usage history. In some embodiments, a personalized knowledge source for a specific user can be extended based on personalized knowledge sources of similar users.

主权项

1. One or more computer-readable media having computer-executable instructions embodied thereon that, when executed by a computing system having a processor and memory, cause the computing system to perform a method for personalizing a language model for a particular target user based on user history information, the method comprising:
receiving usage history information for a target user from one or more sources; parsing the usage history information to identify one or more of an entity or a user action; mapping the one or more of the entity or the user action to a knowledge source, thereby creating a personalized knowledge source; and utilizing the personalized knowledge source to personalize a language model for the target user.