ROMJIST Volume 22, No. 3-4, 2019, pp. 259-271
Hammam M. ABDELAAL, Hassan A. YOUNESS Hadith Classification using Machine Learning Techniques According to its Reliability
ABSTRACT: The Prophetic Hadith is the second source of Islam after the Qur’an, and the fundamental resource of legislation in the Islamic community. This study presents a new methodology to classify the hadith automatically into different categories mainly: Sahih (Sound), Hasan (Good), Da'if (Weak) and Maudu' (Forged) according to the reliability and memory of the reporters (narrators), based on Takhreej al-hadith methods. Hadith authentication depends on three main parts mainly: the Sanad of hadith (chain of narrators), the text, and the references of hadith. The aim of this study is to identify the effective part of them using information gain technique to judge the validity of the hadith, and build a classifier model is able to classify and differentiate hadith classes. Supervised Learning algorithms are used to assign the hadith automatically into one of some predetermined categories, using two classifiers mainly: Decision Trees (DT) and Naive Bayes (NB). The Results showed that DT and NB Classifiers achieved higher accuracy which is reached up to 92.19%, 93.75% respectively, and evaluation of the model using cross-validation is better than percentage split method.KEYWORDS: Data Mining, Text Classification, Machine Learning, Na¨ıve Bayes, Decision Tree, Prophetic HadithRead full text (pdf)