Leveraging Optimization Techniques for Effective Arabic Query Expansion
Abstract
The primary goal of query expansion is to gather terms that are closely associated with the original query terms. To accomplish this goal, similarity measurements are used to evaluate the similarity between the query terms and the collected terms. Unfortunately, existing Arabic query expansion frameworks focus on expanding the query words individually. This approach can affect the expansion process for a complex language such as Arabic. In this paper, we investigate the effectiveness of applying an AI optimization algorithm to Arabic query expansion. The study aims to produce the optimal combination of terms by measuring the similarity of the collected terms against all query terms. We tested two intelligent swarm algorithms: the Cuckoo Search algorithm and Harmony search algorithm. Both algorithms were successful in enhancing the model's precision and recall score with Cuckoo Search algorithm achieving the best performance of (97.5%) precision and (93%) recall.