摘要
传统基于C/S模式的Web搜索方法对网络带宽和网络通畅性的要求都比较高,因此在当今互联网的海量数据中的检索效率比较低。在分析移动Agent技术特点的基础上,提出一种基于移动Agent的分布式Web搜索模型。该模型根据中文信息搜索的特殊性,将移动Agent技术与分类算法相结合,在搜索过程中引入了用户输入信息的预处理过程来进行信息分类,然后根据分类结果形成基于用户兴趣度的移动Agent搜索路径选择策略以及并发方法,由此来提高Web搜索的效率。详细介绍框架的组成和所采用的关键技术,并通过一个移动Agent的开发平台——Aglet平台对模型进行实现和实验。实验结果表明,采用该模型进行Web检索,比传统的C/S检索方式减少50%以上的搜索时间,而在各资源服务器存储的文件资源类别和资源数量差异较大情况下的搜索效率比非并发的其他移动Agent搜索模型搜索时间要减少70%以上。
Traditional Web search methods based on C / S mode have higher requirements in bandwidths and smoothness of network,therefore their retrieval efficiency of mass data in current Internet are rather low. On the basis of analysing the features of mobile Agent technology,we proposed a mobile Agent-based distributed Web search model. According to the particularity of Chinese information search,the model combines the mobile Agent technology with classification algorithms,and introduces the pre-processing process of user inputted information into search process to classify the information,then forms the users interest degree-based route selection strategy of mobile Agent and concurrent method according to classification results,and so as to improve the efficiency of Web search. In the paper we expound the constitution of the framework and the key technologies adopted,as well as the implementation and experiments of the model on Aglet platform,a development platform for mobile Agent. Experimental results indicated that,to carry out Web search using the model could reduce the search time by 50% or lesser than the traditional C/S retrieval mode,and under the condition that there are quite big difference in categories and amounts of documents resources stored in various resource servers,its search efficiency reduced by 70% or higher than the search time of other non-concurrent mobile Agent search model.
出处
《计算机应用与软件》
CSCD
2016年第4期18-21,29,共5页
Computer Applications and Software
基金
广东省省部产学研项目(2012B091100490)
