Intelligent digital newspaper automatic abstract and voice interaction news chat method and system thereof

Abstract

The invention relates to an intelligent digital newspaper automatic abstract and voice interaction news chat method and a system thereof, a neural network is built through a self-attention layer and used for a text intelligent abstract generation method, and the system comprises an algorithm technology, a scheduling technology, an awakening technology, a searching technology, a voice recognition technology, a synthesis technology and a news chat technology, and particularly relates to intelligent abstracts and chat news. The purpose of intelligently and efficiently extracting and summarizing texts from massive news is achieved, and the method is concise, easy to read, personalized in news recommendation and topic recognition and tracking. The real-time pickup, ultrasonic filtering and an attention neural network solve the problems of high dimension, data sparseness and lack of semantic methods of traditional texts, and in voice output, intelligent human-computer interaction experienceof selection, listening, speaking, understanding and answering is given.

Description

technical field [0001] The invention belongs to the field of automatic generation technology of intelligent news processing summaries, natural language processing technology and voice interactive chat news, and in particular relates to an automatic generation method of intelligent news summaries based on semantic correlation, a voice interaction processing method and a system. Background technique [0002] Newspapers are one of the oldest means in the field of communication. For thousands of years, newspapers have made outstanding contributions to the progress of human civilization. As an important channel for news readers to obtain social information resources, news information is an important means for people to obtain effective social information in the era of big data. How to quickly obtain the information we need from this huge ocean of information has become crucial. Intelligent news summary refers to one or a paragraph of sentences extracted and summarized from the so...

AI Technical Summary

Problems solved by technology

[0008] In-depth learning method, voice interaction, learning to obtain dense document quantification of news texts, processing doc2vec word vector representation, so as to solve the problems of high dimensionality, data sparseness, lack of semantics, sentence segmentation and other problems of traditional text representation, using news text word vectors as input, through The attention neural network automatically learns and extracts news text features, thereby avoiding the shortcomings of time-consuming, labor-intensive and error accumulation of traditional news text classification methods that manually participate in feature extraction

Images