Abstract (EN):
A critical component in the implementation of an efficient tabling system is the design of the data structures and algorithms to access and manipulate tabled data. Arguably, the most successful data structure for tabling is tries. However, when used in applications that pose many queries and/or have a large number of answers, tabling call build arbitrarily many and/or very large tables, quickly filling Lip memory. In this paper, we propose a new design for the table space organization where all terms in tabled subgoal calls and tabled answers are represented only once in a common global trie instead of being spread over several different trie data structures. Our initial experiments using the YapTab tabling system show significant reductions oil memory usage without compromising running time.
Language:
English
Type (Professor's evaluation):
Scientific
Contact:
jcosta@dcc.fc.up.pt; jraimundo@dcc.fc.up.pt; ricroc@dcc.fc.up.pt
No. of pages:
15