Abstract:
The Internet is a vital source to disseminate and share information to the masses. This has made information available in abundance on the Web. However, finding relevant information is difficult if not impossible. This difficulty is bilateral between information providers and seekers in terms of information presentation and accessibility respectively.
The primary objective of the online P2P expertise matching system is to build a decentralized Internet scale distributed system for matching experts on the Web. The system is aimed to be highly scalable, self-organizing and robust way of matching experts, Moreover access to information only is not sufficient, as individuals often need to communicate with each other in order to find more important information for solving problems that cannot be obtained from an explicit documentation.
This project focuses on the design and implementation of a peer to peer architecture for an expert matching system. The main functionalities of this project are to create a net peer group, send peer advertisement on the p2p network when a new peer joins that peer group. Peer discovery is made to get the active peers on the network. If an expert matching query is not satisfied by a particular peer, the query is propagated to the other active peers on the network. Other peers may be running on different platforms or may be using different RDBMS. The complex queries to heterogeneous databases are catered by the database manager layer. The results from different peers are returned to the sender peer database manager; the results are merged and returned to the java servlet engine.
