Amazon Technologies, Inc.

Mallory McManamon

Rahul Gupta

Spyridon Matsoukas

Peter Van Tuyl Bentley

Benjamin Lee Shepherd

Chengwei Su

Sankaranarayanan Ananthakrishnan

Shirin Saleem

Chungnam Chan

Joe Virgil Fernandez

Vijayabaskar Gangadaran

Luca Soldaini

Shailesh Srivastava

Yugang Li

Techniques for processing a user input are described. Text data representing a user input is processed with respect to at least one finite state transducer (FST) to generate at least one FST hypothesis. Context information may be required to traverse one or more paths of the at least one FST. The text data is also processed using at least one statistical model (e.g., perform intent classification, named entity recognition, and/or domain classification processing) to generate at least one statistical model hypothesis. The at least one FST hypothesis and the at least one statistical model hypothesis are input to a reranker that determines a most likely interpretation of the user input.

Natural language understanding processing

Speech-controlled searching and browsing for content using speech-controlled devices, or other input-limited devices, is described. A user may audibly indicate to a speech-controlled device whether the user wants to search or browse for content, along with a topic of the content/results to be retrieved. A server, located remotely from the speech-controlled device determines an appropriate endpoint device for displaying results of the requested search or browse. The server also determines an appropriate content source for the requested content, and sends a request for the content to the content source. The server receives search or browse results from the content source and forwards them to the determined endpoint device.

Output of content based on speech-based searching and browsing requests

Growing community of inventors

Seattle, WA, United States of America

Mallory McManamon