一种基于知识的自适应信息过滤技术

Knowledge-based adaptive message filtering technique

提出了一种适用于协同虚拟环境系统的信息过滤技术.分析了不同类型协同工作的协同要求及传统信息过滤技术的不足,给出四种协同工作类型的定义和利用协同标志性物体来判断用户所从事协同工作类型的方法,并介绍了建立协同知识库的过程.系统根据协同标志性物体判断用户当前的协同工作类型,从协同知识库中获取阈值并对DR(deadreckoning)算法进行调整.实验结果表明,该技术不仅满足不同类型协同工作的协同要求,而且显示了较好的信息过滤性能.

Based on the phenomenon that different types of collaborative work may present different request to CVE (collaborative virtual environment) systems, a message filtering technique was proposed. The collaborative work's request for message transmission delay and user behavior simulation accuracy was analyzed, and the reason why traditional message filtering technique could not be used in CVE systems was given. The definition and the judging method of four types of collaborative work were described. Collaborative symbolic entity was introduced to identify the type of collaborative work. System judges the type of work being done by user, then gets the threshold value from collaborative knowledge base and adjusts the DR algorithm with this value. Experimental results show that this technique satisfies not only collaboration request but also system's scalability.