Local search for string problems: Brute force is essentially optimal

Jiong Guo, Dan Hermelin, Christian Komusiewicz

Research output: Contribution to journalConference article


We address the problem of whether the brute-force procedure for the local improvement step in a local search algorithm can be substantially improved when applied to classical NP-hard string problems. We examine four problems in this domain: Closest String, Longest Common Subsequence, Shortest Common Supersequence, and Shortest Common Superstring. Herein, we consider arguably the most fundamental string distance measure, namely the Hamming distance, which has been applied in practical local search implementations for string problems. Our results indicate that for all four problems, the brute-force algorithm is essentially optimal.

Original languageEnglish
Pages (from-to)130-141
Number of pages12
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
StatePublished - 24 Sep 2013
Event24th Annual Symposium on Combinatorial Pattern Matching, CPM 2013 - Bad Herrenalb, Germany
Duration: 17 Jun 201319 Jun 2013


Dive into the research topics of 'Local search for string problems: Brute force is essentially optimal'. Together they form a unique fingerprint.

Cite this