A community is sub-network inside P2P networks that partition the network into groups of similar peers to improve performance by reducing network traffic and high search query success rate. Large communities are common in online social networks than traditional file-sharing P2P networks because many people capture huge amounts of data through their lives. This increases the number of hosts bearing similar data in the network and hence increases the size of communities. This article presents a Memory Thread-based Communities for our Entity-based social P2P network that partition the network into groups of peers sharing data belonging to an entity–person, place, object or interest, having its own digital memory or be a part another memory. These connected peers having further similarities by organizing the network using linear orderings. A Memory-Thread is the collection of digital memories having a common reference key and organized according to some form of correlation. The simulation results show an increase in network performance for the proposed scheme along with a decrease in network overhead and higher query success rate compared to other similar schemes. The network maintains its performance even while the network traffic and size increase.

中文翻译：

---

构建社区以在社交P2P网络中共享人类数字记忆

社区是P2P网络内部的子网，将网络划分为相似对等体的组，以通过减少网络流量和提高搜索查询成功率来提高性能。与传统的文件共享P2P网络相比，大型社区在在线社交网络中更为常见，因为许多人一生中都会捕获大量数据。这增加了网络中承载相似数据的主机数量，因此增加了社区规模。本文为我们基于实体的社交P2P网络提供了一个基于内存线程的社区，该社区将网络划分为共享数据的对等体组，这些数据属于一个实体，一个人，一个地方，一个对象或一个兴趣，拥有自己的数字内存或属于另一部分记忆。通过使用线性顺序组织网络，这些连接的对等方具有进一步的相似性。内存线程是具有公共参考密钥并根据某种相关形式进行组织的数字内存的集合。仿真结果表明，与其他类似方案相比，该方案的网络性能有所提高，同时网络开销减少，查询成功率更高。即使网络流量和大小增加，网络也能保持其性能。