Evaluation of Quranic text retrieval system based on manually indexed topics

This paper investigates the effectiveness of a state of the art information retrieval (IR) system in the verse retrieval problem for Quranic text. The evaluation is based on manually indexed topics of the Quran that provides both the queries and the relevance judgments. Furthermore, the system is evaluated in both Malay and English environment. The performance of the system is measured based on the MAP, the precision at 1, 5 and 10, and the MRR scores. The results of the evaluation are promising, showing the IR system has many potential for the Quranic text retrieval.