TY - GEN
T1 - ACP2P
T2 - 4th International Workshop on Agents and Peer-to-Peer Computing, AP2PC 2005
AU - Mine, Tsunenori
AU - Kogo, Akihiro
AU - Amamiya, Makoto
PY - 2006
Y1 - 2006
N2 - The Agent-Community-based Peer-to-Peer Information Retrieval (ACP2P) method[1],[2] uses agent communities to manage and look up information of interest to users. An agent works as a delegate of its user and searches for information that the user wants by communicating with other agents. The communication between agents is carried out in a peer-to-peer computing architecture. Retrieving information relevant to a user query is performed with content files which consist of original and retrieved documents, and two histories: a query/retrieved document history and a query/sender agent history. The ACP2P is implemented using the Multi-Agent Kodama framework. In this paper, we present some mathematical aspects of the ACP2P method with respect to the relationships between communication loads and the number of records that are stored both in the two histories and retrieved document content files, and discuss the experimental results, for which illustrate the validity of this approach. The results confirm the mathematical conjectures we presented and show that the two histories are more useful for reducing the communication load than a naive method employing 'multicast' techniques, and lead to a higher retrieval accuracy than the naive method.
AB - The Agent-Community-based Peer-to-Peer Information Retrieval (ACP2P) method[1],[2] uses agent communities to manage and look up information of interest to users. An agent works as a delegate of its user and searches for information that the user wants by communicating with other agents. The communication between agents is carried out in a peer-to-peer computing architecture. Retrieving information relevant to a user query is performed with content files which consist of original and retrieved documents, and two histories: a query/retrieved document history and a query/sender agent history. The ACP2P is implemented using the Multi-Agent Kodama framework. In this paper, we present some mathematical aspects of the ACP2P method with respect to the relationships between communication loads and the number of records that are stored both in the two histories and retrieved document content files, and discuss the experimental results, for which illustrate the validity of this approach. The results confirm the mathematical conjectures we presented and show that the two histories are more useful for reducing the communication load than a naive method employing 'multicast' techniques, and lead to a higher retrieval accuracy than the naive method.
UR - http://www.scopus.com/inward/record.url?scp=84884645896&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84884645896&partnerID=8YFLogxK
U2 - 10.1007/11925941_12
DO - 10.1007/11925941_12
M3 - Conference contribution
AN - SCOPUS:84884645896
SN - 3540490256
SN - 9783540490258
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 145
EP - 158
BT - Agents and Peer-to-Peer Computing - 4th International Workshop, AP2PC 2005, Revised Papers
Y2 - 25 July 2005 through 25 July 2005
ER -