Lempel-Ziv-based Hyper-Heuristic Solution for Longest Common Subsequence Problem

The longest common subsequence of a given set of strings is a string whose characters are common among all of them with two properties of being ordered but not consecutive. The general form of LCS problem belongs to the NP-hard problems, therefore, finding the exact LCS cannot be done in a reasonable time. Thus, heuristic methods have been proposed to find the best possible LCS. However, each heuristic is useful for a specific type of set. Some of them are useful for highly correlated strings and others are useful for less-correlated sets. Therefore, there is no single heuristic for any type of the set. This work introduces a hyper-heuristic that computes the correlation among a given set and accordingly assigns the given set to the proper heuristic. To compute a correlation measure, it uses the concepts of the Lempel-Ziv complexity and the longest common substring. The results show that the proposed hyper-heuristic outperforms the state-of-the-art heuristics and hyper-heuristics.