An Algorithm for Arabic Semantic Matching Using New WordNet Tree

The field of Arabic Semantic Matching research has emerged as a critical domain, transforming our comprehension and examination of Arabic language data. In this paper an algorithm to measure the semantic similarity between two Arabic text is proposed based on a new Arabic WorldNet tree, the suggested WordNet is populated and tested using Quranic words and verses. We used Quran corpus to develop the new Arabic model for semantic matching due to the availability of the Quranic words dictionaries and there are several studies handling the semantic matching. Therefore, we can compare the results. After the suggested WordNet is introduced, a new formula to calculate the semantic similarity is suggested and an application is developed. The results of random queries show that the average precision is about 85%.