Attribute matching method in knowledge base questions and answers based on neural network

Abstract

An attribute matching method in knowledge base questions and answers based on a neural network comprises the following steps of replacing entities in questions, generating candidate attributes according to the entities, and sending the segmented words of the candidate attributes to a word embedding layer in the neural network; learning semantic representation of the upper and lower questions by using bidirectional LSTM; calculating according to the word vector representation of the questions and attributes to obtain a word meaning similarity matrix, and similarly, obtaining a semantic similarity matrix according to semantic representation; taking the maximum values of the two similarity matrixes from the question direction and the attribute direction respectively to obtain four vectors, and then obtaining the similarity of the question and the attribute through a full connection layer; and selecting the highest similarity and the attribute corresponding to the highest similarity, if the similarity is greater than a threshold, adding the attribute, and replacing the text in the question to perform the next round of attribute matching. According to the method, the final similarity iscalculated by considering the context semantic representation and the word meaning representation of the questions and the attributes, so that the attribute matching accuracy is improved; corresponding predicate texts in the questions can be positioned, and the multi-hop problem can be iteratively processed.

Description

technical field [0001] The invention belongs to the field of computer technology, relates to an attribute matching technology in knowledge base question answering, and is an attribute matching method in knowledge base question answering based on a neural network model. [0002] technical background [0003] With the rapid development of the information society, massive amounts of data are generated every day, and how people can obtain the required information from large amounts of data is a difficult problem. The question-and-answer system of the knowledge base was born in response to the times, representing a large amount of data as a ternary relationship group with attributes as a bridge between entities, thereby constructing a knowledge map, and establishing a question-and-answer system based on the knowledge map to support user interaction, allowing users to obtain accurate and concise answers . The main work of the knowledge base question answering system is to understa...

AI Technical Summary

Problems solved by technology

The traditional method is easy to explain, can get the corresponding position of knowledge base attributes and question predicates, and can deal with the multi-hop problem by replacing the identified entities and attributes in the question until the next attribute recognition can not be recognized until the new attribute cannot be recognized; A large amount of external information needs to be utilized, templates are manually constructed, and domain migration is limited

Therefore, attribute matching methods based on neural network models have been proposed in recent years, and some models can only handle a limited number of attribute matching. , combining convolutional networks and recurrent networks, but can only handle single-attribute matching; Publication No. CN109271506A "A Method for Constructing a Knowledge Graph Question Answering System in the Field of Power Communication Based on Deep Learning" proposes multiple convolutional neural networks MCCNNs, limited to attribute matching within two hops

In addition, although some models can deal with attribute matching of multi-hop problems, they often require more accurate grammatical structure information. These models are currently mostly used in the English field, and the effect is poor, and they cannot be applied to word segmentation problems and more Matching Chinese attributes for complex grammatical structures

Images