An intelligent question answering system for deep learning in the tax field

Abstract

The invention mainly adopts ansj, hanlp tools, AIML technology and depth learning technology to apply to intelligent question answering, and constructs an intelligent question answering system appliedto tax field. The basic database of the system is established by using crawler to obtain the data set of tax consulting and answering pairs on the Internet, and combining with the data set of 12366 service hotline in the tax bureau and the question answer equivalence extracted from the relevant laws and regulations, and then the dictionary database of word segmentation for deep learning is constructed. Based on the basic question answering database, a model based on template matching is constructed. Based on 12366 question answering database, a retrieval-based model is constructed. Based on the dictionary library, the deep learning model is constructed, which is based on context information and problem types. Based on the template matching model and deep learning model, an intelligent question answering system for tax field is constructed. The invention realizes the automatic recommendation of relevant tax inquiry questions in combination with the main complaint information of the user, and the active question-answer interaction is carried out.

Description

technical field [0001] The invention relates to a natural language processing technology, in particular to a deep learning intelligent question answering system applied in the field of taxation. Background technique [0002] The deep learning intelligent question answering system applied in the tax field is mainly based on natural language processing and AIML technology to build an intelligent question answering system applied in the tax field. [0003] In recent years, the intelligent question answering system has made great development and progress, and many intelligent question answering system products have come out. For example, Watson, an intelligent question answering robot developed by IBM, defeated human contestants in an American quiz show. Apple's Siri system and Microsoft's cortana have achieved good results in the iPhone and Windows 10 operating systems respectively. In China, many enterprises and research groups have also launched many robots with intelligent ...

AI Technical Summary

Problems solved by technology

The method based on deep learning is to learn an answer generation model based on historical question and answer pairs, and generate corresponding answers to user questions. This method has the advantage of simple and easy to expand, but the disadvantage is that the quality of the generated answers is not high and needs to be improved.

Images