摘要
针对用户意图在主动服务各个环节中的不同作用,采用混合智能算法在主动服务中对用户意图进行辨识。用v-SVC的支持向量学习算法对用户的偏好和意图进行学习和辨识,反映用户的状态和所处的角色;用概念层次生成算法对用户服务属性层次化,以获得用户的服务焦点;使用基于模糊聚类算法辨识用户所期望的服务质量或等级。实验分析表明,使用混合算法辨识用户意图具有高效性和实用性。
Based on the different activities of user intention in Web initiative services, a hybrid intelligent algorithm to discover and distinguish user intention was presented. The preference learning with v-SVC provided effective preference recognition and role selection for user intention. The concept hierarchy algorithm was used to conceptualize the user service attributes, thus to find the focus point of service. Fuzzy clustering method was used to identify the service quality and grade expected by the users. The experiment results show that hybrid mining is effective and applicable in intention recognition.
出处
《吉林大学学报(工学版)》
EI
CAS
CSCD
北大核心
2007年第2期419-423,共5页
Journal of Jilin University:Engineering and Technology Edition
基金
国防重大基础预研项目(S0500A001)
关键词
计算机应用
WEB服务
意图识别
主动服务
支持向量机
概念层次
模糊聚类
computer application
Web service
intention recognition~ initiative service
supportedvector machine
concept hierarchy
fuzzy clustering