ArgP2P: An argumentative approach for intelligent query solving in P2P networks

NICOLINI, ANA; ANA MAGUITMAN; CARLOS IVÁN CHESÑEVAR

Buenos Aires

Workshop; TAFA Intl. Workshop on Argumentation; 2015

IJCAI

The Internet is a cooperative and decentralized network builtout of millions of participants that store and share large amounts ofinformation with other users. Peer-to-peer (P2P) systems go hand-in-hand with this huge decentralized network, where each individual nodecan serve content as well as request it. In this scenario, thematic searchalgorithms should lead to and benet from the emergence of semanticcommunities that are the result of the interaction among participants.As a result, intelligent algorithms for neighbor selection should give riseto a logical network topology reecting ecient communication patterns.When routing queries within a P2P network dierent conicting issuesmay arise in individual nodes, such as deciding whether to propagatea query or to reject its processing. Such issues emerge in the contextof incomplete and potentially inconsistent information in a distributedsetting. To the best of our knowledge, current algorithmic approachesto P2P query processing are mostly based on a lack box" approach,providing intelligence to the individual nodes in terms of hard-wiredcode.This paper presents a novel approach to use argumentation as part ofthe decision making machinery within individual nodes in a P2P networkfor thematic search. Our approach will be based on assumption-basedargumentation (ABA). We provide a formalization for P2P networks forthematic search, on top of which intelligent algorithms based on ABAare specied. A case study is used to illustrate the proposed approach,providing as well empirical evaluation of its performance.