Enabling Search and Similarity Search in Small-World-based P2P Systems

Recently, peer-to-peer systems have become one of the most popular distributed applications. Many previous works have investigated identifier-based indexing systems that support a query-by-identifier service. However, clients usually have only partial information about an object, and prefer to query by keywords. In this paper, we propose a small-world-based keyword search system (SW-KSS) that provides keyword search and similarity search services simultaneously. The proposed SW-KSS applies the concept of the "small world theory" to the construction of an indexing structure. Such structures mirror the way humans keep track of their friends and acquaintances; hence, they can cluster peers who share common interests. The method enables a peer to And objects of interest from similar neighboring peers efficiently. We evaluate the performance of SW-KSS via simulations. The results show that SW-KSS can achieve both scalability and partial-match look-up capability.