Keyword search in unstructured peer-to-peer networks

Abstract

Keyword search is a preliminary application for peer-to-peer (P2P) networks. It is important for users to find relevant resources in the highly dynamic system. Many factors affect the results of keyword search, including the underlying structure of peer-to-peer network, the dynamics of peers, the distribution of resources, etc. This chapter introduces the methods used for keyword search in unstructured peer-to-peer networks, and further discusses their extensions for multi-keyword search. These methods can be categorized into two types. One is blind routing. Methods of this type do not consider the distribution of resources. Hence, they get the name of "blind routing". These methods are typically robust. However, their network traffics are high. The other is routing indices. Methods of this type exploit the distribution of resources or query keywords. Therefore, they have low network traffic, especially for popular resources or queries. For each method, an algorithm flow is presented, followed by an analysis of the pros and cons. A comparison is also made to demonstrate the differences between these methods. At the end of this chapter, a discussion on extending the search methods to the multi-keyword search problem is held. The methods introduced in this chapter work differently in the multi-keyword search scenario. Some may need no adaption while some shall be modified for multi-keyword indices. The efficiencies of these methods in this problem are also considered and compared. © 2010 Springer Science+Business Media, LLC.