Title :
Network-based simple recursive answer evaluation for deductive databases in parallel environment
Author :
Kim, Kunsu ; Henschen, L.J.
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Northwestern Univ., Evanston, IL
Abstract :
Because of increasing needs and requirements for the use of databases, one always looks for more efficient ways to handle database access. Parallel computing environments draw more attention to achieve the high processing speed and the less expensive processing method. This paper proposes the possibility of using a connectionist model by treating every datum as an active processing unit cooperating with other such units via messages in getting answers in a deductive database, especially with recursion. It uses query compilation and iteration to process the recursive case. The notion of the reverse-compilation which is essential to get the correct answers is introduced in handling recursive cases. It introduces the recursive controller which is dedicated to processing the recursive cases as a sublayer to a central database controller
Keywords :
deductive databases; parallel processing; query processing; recurrent neural nets; connectionist model; database access; deductive database; messages; parallel environment; query compilation; query iteration; recurrent neural nets; recursion; recursive controller; reverse-compilation; Centralized control; Deductive databases; Parallel processing;
Conference_Titel :
Software Engineering and Knowledge Engineering, 1992. Proceedings., Fourth International Conference on
Conference_Location :
Capri
Print_ISBN :
0-8186-2830-8
DOI :
10.1109/SEKE.1992.227946