Title :
Research of tourist attractions hotel QA system based on short message
Author :
Mao Cun-li ; Yu Zheng-tao ; Xian Yan-Tuan ; Guo Jian-yi ; Lei Xiong-Li
Author_Institution :
Sch. of Inf. Eng. & Autom., Kunming Univ. of Sci. & Technol., Kunming, China
Abstract :
To the feature of service database of tourist attraction and hotel in tourism field, system framework and implementation of tourist attraction, hotel QA system based on short message is described, in which attraction and hotel information is stored in database. Attraction and hotel are defined as domain entity, entity property, property category and property value. According to the characteristics of short message, domain entity is identified by CRF; query topic (property category of domain entity) is identified by a dynamic assembly classification method for short text classification. In this method, a treelike assembly classifier was constructed to support the text classification. Further, a dynamic adjusting strategy was presented in the construction procedure and inverted index is built by Lucene. We use domain entity and query topic to search the corresponding answer, which is that extracting the property value corresponding to the entity property to achieve QA system of tourist attraction and hotel. Experimental results in Yunnan Tourist show that better result has been achieved; the proposed method increases the accuracy of QA system.
Keywords :
database management systems; hotel industry; pattern classification; query processing; question answering (information retrieval); text analysis; travel industry; trees (mathematics); CRF; Lucene; Yunnan tourist show; construction procedure; domain entity; dynamic adjusting strategy; dynamic assembly classification method; entity property; hotel information; inverted index; property category; property value; query topic; service database; short message; short text classification; system framework; tourism field; tourist attractions hotel QA system; treelike assembly classifier; Assembly; Boosting; Educational institutions; Electronic mail; Indexes; Text categorization; Domain QA; Lucene; Short Message; Topic Classification; Yunnan Tourist;
Conference_Titel :
Control Conference (CCC), 2011 30th Chinese
Conference_Location :
Yantai
Print_ISBN :
978-1-4577-0677-6
Electronic_ISBN :
1934-1768