The study compares the performance of three prominent metaheuristic optimization algorithms namely: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Harmony Search (HS) for solving the fuzzy string-matching problem. In the traditional method the exact matching of the string is been done while in the fuzzy string-matching method we focus on finding the approximate matches by minimizing the character level discrepancies, for this we have used metrics such as Levenshtein distance similarity. The fuzzy approach is mainly used in the domains where the data is faulty, inconsistent or needs to be partially matched similar to natural language processing (NLP) and record linkage. In fuzzy string method we use string with varying lengths and noise levels. Its performance is assessed through convergence analysis, execution time comparison, error metrics such as RMSE and MAE, and statistical hypothesis testing (Wilcoxon signed-rank test and t-test).