2014 IEEE 26th International Conference on Tools with Artificial Intelligence (ICTAI) (2014)
Nov. 10, 2014 to Nov. 12, 2014
Conflict based clause learning is known to be an important component in Modern SAT solving. Because of the exponential blow up of the size of learnt clauses database, maintaining a relevant and polynomially bounded set of learnt clauses is crucial for the efficiency of clause learning based SAT solvers. In this paper, we first compare several criteria for selecting the most relevant learnt clauses with a simple random selection strategy. We then propose new criteria allowing us to select relevant clauses w.r.t. A given search state. Then, we use such strategies as a means to diversify the search in a portfolio based parallel solver. An experimental evaluation comparing the classical Many SAT solver with the one augmented with multiple deletion strategies, shows the interest of such approach.
Databases, Portfolios, Search problems, Space exploration, Hardware, Complexity theory, Data structures
L. Guo, S. Jabbour, J. Lonlac and L. Sais, "Diversification by Clauses Deletion Strategies in Portfolio Parallel SAT Solving," 2014 IEEE 26th International Conference on Tools with Artificial Intelligence (ICTAI), Limassol, Cyprus, 2014, pp. 701-708.