Classification of Hadith Levels Using Data and Text Mining Techniques

Abstract

There is no definite information among ulama, about the beginning of the Prophet's hadith forgery, but this problem has spread and responded to the community. The purpose of the hadith counterfeiters are various motives and motivations, the factors that encourage them to falsify the hadith are to defend certain interests: defending political interests, defending theology, defending fiqh madzhab, attracting people who hear their stories, to dignify others, encourage others are more persistent in worshiping and destroying Islam. Determining the level of hadith requires a long process because we have to read the entire hadith and know the perawi and sanad. This problem requires a solution to overcome it. Through this research, search and analysis of the model was carried out using the Naïve Bayes (NB) algorithm and the Decision Tree algorithm (C4.5). Evaluation is done by comparing two algorithms. Based on the results of the study found that the Decision Tree Algorithm (C4.5) has a higher accuracy rate of 7.81% of the Naïve Bayes Classifier Algorithm.