Constraint-Based Reasoning: An Efficient Problem Solving Strategy in ArtificialIntelligence Applications

摘要

Search problems are widely involved in most applications of artificialintelligence. Finding a subspace which satisfies a set of constraints is the concern of these problems. A constraint-based approach to problem solving has been looked at in artificial intelligence (AI) and other areas of computer science. A variety of methods are subsumed in a technique called constraint satisfaction or reasoning based on relations. Backtracking is a common approach in most of the constraint satisfaction problems. Backtracking suffers from some shortcomings, making it an inefficient approach for properly solving constraint satisfaction problems. A complementary approach for constraint satisfaction problems based on the consistency concept is more efficient than backtracking techniques. The efficiency of these different techniques is a direct consequence of the constraint handling. The constraint concepts and related techniques for solving constraint satisfaction problems are discussed in this paper.