| [Purpose] Based on the analysis of the medical question answering system both at home and abroad;applied a automatic question answering system in the base of semantic technology into Chinese context.The automatic question answering system framework was design to intervene the lifestyle of diabetes to improve the methods of obtaining information via automatic question answering system,and relieve the problem of medical information resources utilization,providing reference for effective use of medical information resources in the future.[Methods] A combination of qualitative and quantitative research methods were used in this study:(1)Based on the researches of automatic question answering system both in theory and practice,a systematic review was conducted to analysis the method and tools which could be used in the Chinese context;(2)Two parts were included in the crawling of Web data,the question experiment set and knowledge base.The question experiment set was obtained by using method of network spider crawling the web data acquisition,knowledge base is used to construct the ontology and rules as knowledge base;(3)The key technologies of the question answering system was used for the experimental analysis,to understand the working process and performance of the automatic question answering system modules.[Results] This study designed a Chinese domain automatic question answering system framework,and tested the key steps in the automatic question answering system.The classification of SVM machine learning algorithm based on Named Entity Recognition in the base of CRF have achieved good effects.[Conclusions] Due to the particularity of Chinese context,the question answering system which is suitable for the English language is not suitable for Chinese environment.The research of this paper has good performance in classification and entity recognition,and it could provide certain reference for the research and application of question answering system. |
PDF Full Text Request