Enhanced question answering interaction system and method for universal mapping knowledge domain on the basis of deep learning

Abstract

The invention discloses an enhanced question answering interaction system and method for a universal mapping knowledge domain on the basis of deep learning. The system comprises an expanded mapping knowledge domain hybrid question answering module, a knowledge base question generation module, a web interaction module, a mapping knowledge domain question answering module and a knowledge depth reasoning module, wherein the expanded mapping knowledge domain hybrid question answering module is used for obtaining an expanded mapping knowledge domain; the knowledge base question generation module independently generates answers corresponding to different questions to enable generate a plurality of question-answer pairs; the web interaction interface is used for obtaining a user question; the mapping knowledge domain question answering module obtains the type of an answer corresponding to the user question, and obtains a numerical value vector corresponding to the user question; and the knowledge depth reasoning module is used for carrying out knowledge retrieval and reasoning on the type of the answer corresponding to the user question and the numerical value vector corresponding to theuser question, and obtaining the target answer of the user question according to retrieval and reasoning results and a plurality of question-answer pairs. By use of the system, the performance, the operability, the semantic comprehension analysis ability, the comprehensive question answering expansion ability and the universal technology sharing ability of the question answering system can be effectively improved, and answer generation accuracy is improved.

Description

technical field [0001] The present invention relates to the technical field of machine learning, in particular to a general knowledge graph enhanced question answering interactive system and method based on deep learning. Background technique [0002] With the development of big data and knowledge engineering, a computing paradigm combining data-driven and knowledge-driven will gradually be formed, and more and more technologies such as information extraction, knowledge fusion and knowledge processing are used to construct massive unstructured data into knowledge. Atlas makes massive data information expressed in a form closer to the human cognitive world, and provides a better ability to organize, manage and understand massive information. The question answering system based on knowledge graph has become a hot research and application direction in academia and industry. At present, question answering systems for knowledge graphs commonly use semantic analysis methods such ...

AI Technical Summary

Problems solved by technology

[0006] However, the main problem in the knowledge graph question answering system of related technologies is that at present, most domain knowledge graph question answering usually adopts template matching, semantic logic expression SPARQL query, complex problem disassembly, simple reasoning and other methods, and these methods consume a lot of resources. Labor costs, and at the level of shallow semantic analysis and simple knowledge reasoning, mainly reflected in the processing of factual problems

Images