公开(公告)号：US20190108257A1

公开(公告)日：2019-04-11

申请号：US15726394

申请日：2017-10-06

Applicant: SoundHound, Inc.

Inventor： Luke Lefebure ,  Pranav Singh

IPC: G06F17/30 ,  G06N7/00 ,  G06F17/27 ,  G10L15/183

Abstract: A speech recognition and natural language understanding system performs insertion, deletion, and replacement edits of tokens at positions with low probabilities according to both a forward and a backward statistical language model (SLM) to produce rewritten token sequences. Multiple rewrites can be produced with scores depending on the probabilities of tokens according to the SLMs. The rewritten token sequences can be parsed according to natural language grammars to produce further weighted scores. Token sequences can be rewritten iteratively using a graph-based search algorithm to find the best rewrite. Mappings of input token sequences to rewritten token sequences can be stored in a cache, and searching for a best rewrite can be bypassed by using cached rewrites when present. Analysis of various initial token sequences that produce the same new rewritten token sequence can be useful to improve natural language grammars.

公开(公告)号：US10599645B2

公开(公告)日：2020-03-24

申请号：US15726394

申请日：2017-10-06

Applicant: SoundHound, Inc.

Inventor： Luke Lefebure ,  Pranav Singh

IPC: G06F17/27 ,  G06F16/2453 ,  G06N7/00 ,  G10L15/183

Abstract: A speech recognition and natural language understanding system performs insertion, deletion, and replacement edits of tokens at positions with low probabilities according to both a forward and a backward statistical language model (SLM) to produce rewritten token sequences. Multiple rewrites can be produced with scores depending on the probabilities of tokens according to the SLMs. The rewritten token sequences can be parsed according to natural language grammars to produce further weighted scores. Token sequences can be rewritten iteratively using a graph-based search algorithm to find the best rewrite. Mappings of input token sequences to rewritten token sequences can be stored in a cache, and searching for a best rewrite can be bypassed by using cached rewrites when present. Analysis of various initial token sequences that produce the same new rewritten token sequence can be useful to improve natural language grammars.