A Question Answering System For Specific Domain

With the increasing development of the Internet, information need by people is also increased rapidly. Nowadays, with the development of intelligent devices, people want a way to interact with computer natually and get the information precisely. This demand is highlighted especially in specific domains(such as healthy care): enterprises and institutions have professional knowledges which many unprofessional people want to know about. So a framework to build a question answering system to provide the knowledge in natural way is in great demand.In this paper, we proposed a general framework for domain-specific interactive question answering systems which takes advance of domain knowledge bases. The framework is consist of several core modules such as semantic paser, dialog manager and bussiness logic module. The semantic parser is responsible for parsing users’ questions to the corresponding logical forms on basis of the knowledge base structure. Then, the logical forms are translated into query language to further harvest answers from the knowledge base.Dialog manager is featured with automatic dialog strategy development which is rely on manual intervention in traditional interactive QA systems. With the dialog manager,user can make general dialog behavior such as further questions. Bussiness logic is due to handle domain-specificed dialog logics as well as answer generation. Moreover, this framework can make use of context informations such as users’ geolocation or personal data to play personalized service. The framework is designed to be domain-independent,therefore it makes the question answering system easy to be transferred to other domains.We applied our framework to a Chinese interactive QA system development, and used a health-related knowledge base as domain scenario. User can get to our website to get diagnosed or ask about normal diseases and medicines. They can also provide their personal data to make the diagose more personalized.