摘要： 共享信息的集中存储对存放这些信息的服务器提出了较高的要求，同时，服务器将成为整个系统的瓶颈。为此，提出了一种基于P2P 的信息共享与推荐模型，解决了信息集中存放产生的问题。接着，对该模型中涉及到的基于内容的过滤，提出了一种基于词汇链的方法，较好地解决了纯粹单一关键词无法准确描述文本的问题，并对信息推荐中使用最成功的协同过滤算法进行了描述。给出了文本过滤的实验结果及其分析。
关键词:
对等网络；客户机/服务器；词汇链；文本过滤；协同过滤
Abstract: To solve the bottleneck of the server and the shortage of reliability about centralizing storage in sharing information system, the distributed information sharing model is put forward, which is based on peer to peer networking. Based on it, the basic theory and the algorithm about content-based documents filtering based on lexical chains are given, and then, the collaborative filtering algorithm is discussed. Finally, the validity of content-based documents filtering algorithm is validated through using the medical corpus OHSUMED on TREC-9.
Key words:
Peer to Peer networking; Client/server; Lexical chain; Document filtering; Collaborative filtering
李绍滋，周昌乐，陈火旺. 基于 P2P 网络的信息过滤与推荐技术研究[J]. 计算机工程, 2006, 32(8): 45-47.
LI Shaozi, ZHOU Changle, CHEN Huowang. Research on Information Filtering and Recommendation Based on Peer to Peer Network[J]. Computer Engineering, 2006, 32(8): 45-47.