Research On Search Strategy For Social Resource In Peer-to-peer Social Networks

As an alternative to the centralized architecture of the traditional Online Social Networks(OSNs),P2 P networks(Peer-to-Peer)can solve the security privacy problems and single point of failure in the centralized architecture.Resource search is a critical component of P2 P social networks.The effectiveness of resource search model design and the efficiency and accuracy of search strategy are important factors that affect the overall performance of P2 P social networks.Therefore,the research of P2 P social network model and resource search strategy is of great practical significance.In this dissertation,the relevant theoretical knowledge and resource search strategy of P2 P network and social networks are studied and introduced,and on this basis,drawing on existing research results,the resource search model and resource search strategy in P2 P social network are studied in detail.The concrete research contents include:(1)The traditional P2 P network resource model is unable to adapt to the dynamic changes of P2 P social networks,resulting in poor resource search efficiency.To solve this problem,an interest-based self-organizing resource search model(IRSM)was designed.Through the use of human social theory,the model allowed users to spontaneously form resource communities and social knowledge networks in a self-organized form,which helped to improve the search efficiency in the future.The entire process consumed no additional network overhead.The blacklist was designed to reduce the impact of network churn which caused by users frequently going online and offline.(2)In view of the problems of traditional resource search strategies,such as large overhead,low efficiency,and inability to support multi-topic complex queries and so on,an adaptive resource search strategy called MTRS(Multi-Topic Resource Search)that supported multi-topic queries was proposed based on the self-organizing resource search model proposed above.Firstly,a strategy query method was proposed to increase the probability of nodes finding the resources in the early stage of network establishment.Secondly,a novel recommendation node acquisition strategy was proposed to find nodes that may contain query resources.Finally,a self-adaptive forwarding algorithm was proposed to adaptively forward the acquired nodes.Adaptive forwarding was used to achieve better search efficiency with less overhead.(3)In this dissertation,a P2 P social network software simulation platform was designed to simulate the resource discovery process in the P2 P network under the dynamic network environment.The comprehensive performance of the resource search strategy was investigated from the recall rate,the number of discovery resources and network overhead.Experimental results showed that compared with other resource search methods,the proposed resource search strategy had higher recall rate and lower network overhead.