After a short discussion of the data structures used in a PROLOG-like inference machine, the paper presents an optimized algorithm for binding search. The software solution was designed for a PROLOG interpreter running on the RSX-11M operating system on PDP-11 compatible computers. It was tested for efficiency and the results are further discussed. The hardware solution is based on a dedicated smart-memory implemented with two building blocks: memory cells (MCs) interconnected one to another as an L-machine and an address register (AR). Estimation of area and time are presented.
|Number of pages||14|
|Journal||Buletinul Institutului Politehnic Bucuresti, Seria Automatica-Calculatore|
|Publication status||Published - 1988|
ASJC Scopus subject areas