At present, how to enable Search Engine to construct user personal interest model initially, master user＇s personalized information timely and provide personalized services accurately have become the hotspot in the r...At present, how to enable Search Engine to construct user personal interest model initially, master user＇s personalized information timely and provide personalized services accurately have become the hotspot in the research of Search Engine area. Aiming at the problems of user model＇s construction and combining techniques of manual customization modeling and automatic analytical modeling, a User Interest Model （UIM） is proposed in the paper. On the basis of it, the corresponding establishment and update algorithms of User lnterest Profile （UIP） are presented subsequently. Simulation tests proved that the UIM proposed and corresponding algorithms could enhance the retrieval precision effectively and have superior adaptability.展开更多
提出了一个基于多Agent的智能信息检索(intelligent information retrieval based on multi-agent,IIR M-Agent)框架。介绍了多Agent群组的工作原理,提出了一个具有二层结构的IIRM-Agent框架,阐述了各个subagent的任务以及它们之间如何...提出了一个基于多Agent的智能信息检索(intelligent information retrieval based on multi-agent,IIR M-Agent)框架。介绍了多Agent群组的工作原理,提出了一个具有二层结构的IIRM-Agent框架,阐述了各个subagent的任务以及它们之间如何协作完成智能检索、自动通告、导航条和个人信息管理的功能。展开更多
为提高基于Agent的信息检索系统在海量的网络信息检索中查询准确率,提出了基于多兴趣Agent层次结构的检索系统模型(IRHOMIA,information retrieval system based on hierarchically organizedofmulti-interest Agent),模型对查询信息进...为提高基于Agent的信息检索系统在海量的网络信息检索中查询准确率,提出了基于多兴趣Agent层次结构的检索系统模型(IRHOMIA,information retrieval system based on hierarchically organizedofmulti-interest Agent),模型对查询信息进行了兴趣预测并生成了用户兴趣项权重向量,输入到训练过的神经网络并把输出层生成向量中的每个值与给定的阈值进行比较来确定将查询任务分配给其他兴趣Agent或者是拥有相应资源的查询工具。试验表明,IRHOMIA在预测用户兴趣的效果和推荐搜索信息的准确率方面比传统的检索系统以及单兴趣Agent检索系统IRHOIA有5%以上的提高。展开更多
随着网络信息资源的迅速增加,如何及时准确地获取所需信息是现代网络信息过滤技术需要解决的主要问题。为了给用户提供更准确的信息,提出了一种基于用户反馈的智能合作过滤模型(Agent collaborative filtering model based on users'...随着网络信息资源的迅速增加,如何及时准确地获取所需信息是现代网络信息过滤技术需要解决的主要问题。为了给用户提供更准确的信息,提出了一种基于用户反馈的智能合作过滤模型(Agent collaborative filtering model based on users' feedback,ACFM)和用户兴趣模型,该模型通过隐式反馈和显式反馈这两种用户兴趣反馈学习实现合作过滤。实验结果表明,ACFM在预测用户兴趣的效果和推荐搜索信息的准确率方面比传统的搜索引擎有明显改善。展开更多
基金Supported by the National Natural Science Foundation of China (50674086)the Doctoral Foundation of Ministry of Education of China (20060290508)the Youth Scientific Research Foundation of CUMT (0D060125)
文摘At present, how to enable Search Engine to construct user personal interest model initially, master user＇s personalized information timely and provide personalized services accurately have become the hotspot in the research of Search Engine area. Aiming at the problems of user model＇s construction and combining techniques of manual customization modeling and automatic analytical modeling, a User Interest Model （UIM） is proposed in the paper. On the basis of it, the corresponding establishment and update algorithms of User lnterest Profile （UIP） are presented subsequently. Simulation tests proved that the UIM proposed and corresponding algorithms could enhance the retrieval precision effectively and have superior adaptability.
文摘提出了一个基于多Agent的智能信息检索(intelligent information retrieval based on multi-agent,IIR M-Agent)框架。介绍了多Agent群组的工作原理,提出了一个具有二层结构的IIRM-Agent框架,阐述了各个subagent的任务以及它们之间如何协作完成智能检索、自动通告、导航条和个人信息管理的功能。
文摘为提高基于Agent的信息检索系统在海量的网络信息检索中查询准确率,提出了基于多兴趣Agent层次结构的检索系统模型(IRHOMIA,information retrieval system based on hierarchically organizedofmulti-interest Agent),模型对查询信息进行了兴趣预测并生成了用户兴趣项权重向量,输入到训练过的神经网络并把输出层生成向量中的每个值与给定的阈值进行比较来确定将查询任务分配给其他兴趣Agent或者是拥有相应资源的查询工具。试验表明,IRHOMIA在预测用户兴趣的效果和推荐搜索信息的准确率方面比传统的检索系统以及单兴趣Agent检索系统IRHOIA有5%以上的提高。
文摘随着网络信息资源的迅速增加,如何及时准确地获取所需信息是现代网络信息过滤技术需要解决的主要问题。为了给用户提供更准确的信息,提出了一种基于用户反馈的智能合作过滤模型(Agent collaborative filtering model based on users' feedback,ACFM)和用户兴趣模型,该模型通过隐式反馈和显式反馈这两种用户兴趣反馈学习实现合作过滤。实验结果表明,ACFM在预测用户兴趣的效果和推荐搜索信息的准确率方面比传统的搜索引擎有明显改善。
