Spoken language semantic analysis system and method

Abstract

The invention discloses a spoken language semantic analysis system and method. The spoken language semantic analysis system is used for carrying out analysis on spoken language semantics in a preset field, and comprises a storage unit, an acquisition unit, an indexing unit and an analysis unit, wherein the storage unit is used for storing semantic sentence patterns in the preset field; each semantic sentence pattern corresponds to one address; the semantic sentence patterns comprise characters and keywords; each keyword corresponds to one label; a word list is preset in the storage unit and is used for storing the address of the semantic sentence pattern in which each character is positioned and / or the address of the semantic sentence pattern in which each label is positioned; the acquisition unit is used for acquiring a spoken language sentence to be analyzed; the indexing unit is respectively connected with the storage unit and the acquisition unit and is used for carrying out retrieval on the semantic sentence patterns in the storage unit according to the spoken language sentence to be analyzed and acquiring candidate semantic sentence patterns which accord with the spoken language sentence to be analyzed and a corresponding candidate sequence; and the analysis unit is connected with the indexing unit and is used for carrying out analysis on the spoken language sentence to be analyzed according to the ranked candidate semantic sentence patterns by adopting a fuzzy matching algorithm and obtaining an analysis result.

Description

technical field [0001] The invention relates to the field of natural language understanding of spoken language, in particular to a highly robust spoken language semantic analysis system and method. Background technique [0002] Spoken speech recognition involves multidisciplinary fields such as phonetics, linguistics, mathematical signal processing, and pattern recognition. With the popularity of smart devices, how to interact more directly and friendly between people and smart devices has become an important issue. Due to the natural friendliness and convenience of spoken natural language for users, human-computer interaction based on spoken natural language has become a trend and has attracted more and more attention from the industry. The key technology of spoken natural language interaction lies in the semantic understanding of spoken language, that is, to analyze the user's spoken sentences to obtain the intention and corresponding keywords that the user wants to expre...

AI Technical Summary

Problems solved by technology

In the existing semantic analysis methods of spoken language, most of them are based on certain grammar matching, such as regular grammar and context-free grammar, which require that the spoken sentence to be parsed must be completely consistent with the semantic sentence structure in order to successfully parse; this makes the semantic understanding system The constructors need to spend a lot of time collecting semantic sentence patterns; due to the inaccurate recognition of front-end speech recognition and other modules, the parsing of semantic understanding fails; and because the sentence patterns to be parsed need to be matched with a large number of semantic sentence patterns, It will cause the problem of long analysis time and low efficiency

Images