Arabic Word Sense Disambiguation for Information Retrieval
Mohammed Alaeddine Abderrahim, Mohammed El-Amine Abderrahim · ACM Transactions on Asian and Low-Resource Language Information Processing · 2022
In the context of using semantic resources for information retrieval, the relationship and distance between concepts are considered important for word sense disambiguation. In this article, we experiment with Conceptual Density and Random Walk with graph methods to enhance the performance of the Arabic Information Retrieval System. To do this, a medium-sized corpus was used. The results proved that Random Walk can enhance the performance of the information retrieval system by achieving a mean improvement of 13%, 16%, and 12% in terms of recall, precision, and F-score, respectively.