Automatic recognition and extraction method for problematic target features

Abstract

The invention discloses a problem target feature automatic identification and extraction method. The method comprises the following steps: generating at least one dependency relationship chain sampleaccording to a preset generation rule; according to a preset matching rule, matching a candidate dependency relationship chain matched with the to-be-analyzed question text from the at least one dependency relationship sample; and according to a preset screening rule, screening problem target words of the to-be-analyzed problem text from the candidate dependency relationship chain. The question target word recognition method has the advantages that the words related to the question intention of the user are recognized from the to-be-analyzed text through the question target word recognition method, and therefore the intention of the user can be recognized more accurately.

Description

technical field [0001] The embodiment of the invention relates to the field of computer technology, in particular to a method for automatic identification and extraction of problem target features. Background technique [0002] At present, with the development of artificial intelligence technology and big data technology, question answering system, as an advanced application form of information retrieval system, generally exists in various research fields such as professional services, education, and life. Among them, the accurate identification and classification of the target of the question raised by the user has a direct impact on the question answering system to accurately identify the user's information needs, filter candidate answers, and improve user satisfaction with the answer. [0003] However, the candidate answers provided by the existing question answering system are not accurate enough. The main reason is that the existing question answering system identifies ...

AI Technical Summary

Problems solved by technology

[0003] However, the candidate answers provided by the existing question answering system are not accurate enough. The main reason is that the existing question answering system identifies and classifies the subject of the question by screening the subject words of the user's question, while ignoring the importance of the identification and classification of the question target.

The identification and classification of question topics and question targets are not the same. Question topics focus on describing the main object of the question content, while question objectives focus on describing the types of answers users expect. The question answering system of recognition and classification cannot give precise answers to the intention of the user to ask the question

Images