TY - GEN
T1 - Answering cooperative recursive queries in web federated databases
AU - Balaban, Mira
AU - Berezansky, Nikolai
AU - Gudes, Ehud
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2002.
PY - 2002/1/1
Y1 - 2002/1/1
N2 - Evaluation of recursive queries and computing transitive closures require multiple accesses to the involved relations. In a federated database this leads to multiple accesses to the participants of the federation. Since the components are not uniform in terms of computation power, reliability, and communication delays, it might be desirable to minimize the number of accesses to the individual databases, and to maximize the size of the obtained answer with respect to time. Based on this observation, we developed cooperative query planning methods, termed Deep Federated Semi-Naive (DFSN), for computing the strong partial transitive closure of a relation. We have implemented and tested these algorithms in a real database environment. The experimental results show better performance of the DFSN methods over the conservative semi-naive approaches in that they produce large answer sets in time that is considerately shorter than the time needed by the conservative approaches.
AB - Evaluation of recursive queries and computing transitive closures require multiple accesses to the involved relations. In a federated database this leads to multiple accesses to the participants of the federation. Since the components are not uniform in terms of computation power, reliability, and communication delays, it might be desirable to minimize the number of accesses to the individual databases, and to maximize the size of the obtained answer with respect to time. Based on this observation, we developed cooperative query planning methods, termed Deep Federated Semi-Naive (DFSN), for computing the strong partial transitive closure of a relation. We have implemented and tested these algorithms in a real database environment. The experimental results show better performance of the DFSN methods over the conservative semi-naive approaches in that they produce large answer sets in time that is considerately shorter than the time needed by the conservative approaches.
UR - http://www.scopus.com/inward/record.url?scp=84937398306&partnerID=8YFLogxK
U2 - 10.1007/3-540-45431-4_3
DO - 10.1007/3-540-45431-4_3
M3 - Conference contribution
AN - SCOPUS:84937398306
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 14
EP - 28
BT - Next Generation Information Technologies and Systems - 5th International Workshop, NGITS 2002, Proceedings
A2 - Halevy, Alon
A2 - Gal, Avigdor
PB - Springer Verlag
T2 - 5th International Workshop on Next Generation Information Technologies and Systems, NGITS 2002
Y2 - 24 June 2002 through 25 June 2002
ER -