摘要
提出一种软定时转变自动机,识别更大范围的用户行为历史。该软建模的框架用户模型可以在类似网络学习的虚拟平台中实时地评估用户的动态行为,通过评估观察到的行为中的时间偏差、额外的或省去的行为量来评估用户行为。当观察到的行为只部分满足给定的行为模型约束时,该模型也可以用于用户历史行为的部分识别。相对于标准的布尔模型的识别,当多个用户模型存在偏差量时,该方法通过预测、投影和其他已知的技术更适用于用户实时支持系统,引导生成系统支持。基于日志的网络学习平台和软时间自动机的规划实验证明了该模型的表现力和有效性。
A soft timed transition automata(TTA)is proposed to recognize the user history behavior in a wider range.The user behavior model constructed with soft modeling can evaluate the dynamic behavior of users in real time in the virtual platform similar to network learning.The time deviation,additional or omitted behavior amount observed in estimation are used to evaluate the user behavior.When the observed behavior partially satisfies the given constraint of behavior model,the model can be used to the part recognition of user history behavior.In comparison with the standard Boole model recognition,the proposed method is more suitable for the user real-time support system by means of prediction,projection and other know techniques while the deviation exists in a large number of user model,and can be used to generate the system support.The expressivity and validity of the model were verified with planning experiment of network learning platform based on logs and soft timed automata.
作者
许小媛
纪河
黄黎
李从明
XU Xiaoyuan;JI He;HUANG Li;LI Congming(College of Information and Mechanical Engineering,Jiangsu Open University,Nanjing 210017,China;College of Computer Science and Technology,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《现代电子技术》
北大核心
2018年第15期165-168,共4页
Modern Electronics Technique
基金
江苏省教育科学"十三五"规划2016年度青年专项课题(C-b/2016/03/25)
江苏开放大学"十三五"2016年度重点课题(16SSW-Z-003)
江苏省高校自然科学基金项目(15KJB520005)
安徽省高校自然科学重点项目(KJ2015A366)~~
关键词
用户行为
时变自动机
自动规划
建模
软定时
匹配
user behavior
time varying automaton
automated planning
modeling
soft timing
matching
