Voice interaction realizing method

Abstract

The invention discloses a voice interaction realizing method. The voice interaction realizing method includes: step 1, establishing a semantic recognition node library; step 2, loading the semantic recognition node library; step 3, processing a voice signal by subjecting an input voice signal to data processing so as to form multiple literal strings; step 4, matching with the semantic recognition node library: enabling the literal strings to match with the semantic recognition node library, turning a return result from the literal strings into a corresponding node ID (identifier) on success, or processing as an error and finishing on failure; step 5, matching with a voice library successfully, performing voice broadcast according to the return result of the corresponding node ID, and finishing a voice interaction process after broadcasting is over. The voice interaction realizing method has the advantages that multiple enquiring manners for one same semanteme are integrated into one node, an answer can be matched directly during voice recognition, and a one-stop voice reaction process integrating voice recognition, semantic analysis and answer matching is formed.

Description

Technical field: [0001] The invention relates to a method for implementing voice interaction, and relates to the technical field of artificial intelligence. Background technique: [0002] Speech recognition technology, also known as Automatic Speech Recognition, (ASR), whose goal is to convert the lexical content in human speech into computer-readable input, 1920, a toy dog ​​named Radio Rex It should be considered the earliest speech recognizer, because the dog can respond with actions such as "popping off the base" when the dog's name is called. The Audrey speech recognition system developed by AT&T Bell Laboratories in 1952 can recognize 10 English numbers. In the 1960s, artificial neural networks were introduced to speech recognition. The two breakthroughs in this era are Linear Predictive Coding (LPC) and Dynamic Time Warp technology. After the 1980s, the focus of research gradually shifted to large vocabulary, non-specific continuous speech recognition. There have ...

AI Technical Summary

Problems solved by technology

In the 1960s, foreigners conducted a large-scale research work on machine translation and spent a lot of money, but the final result was unsatisfactory. People miscalculated the complexity of natural language, and the processing technology and theory of the time They are still immature, so it is taken for granted that there will be no development prospects

Natural language analysis is also called natural language processing. At present, many processing methods have been developed for multiple languages. The main forms are word segmentation and semantic analysis. The whole process of voice interaction is simply explained by voice recognition into text strings, and then the text strings are processed. Word segmentation and semantic analysis, and then find the answer through keyword matching for broadcasting, but the speech recognition and natural language analysis of this voice interaction method are carried out step by step, and the interaction process is cumbersome and the accuracy and intelligence are not high.

Images