Hadith Authenticity Prediction using Sentiment Analysis and Machine Learning

Abstract

Starting around 815AD/200AH scholars have put immense effort towards gathering and sifting authentic hadiths, which are prophetic traditions of the Muslim community. The authenticity of a hadith solely depends on the reliability of its reporters and narrators. Till now scholars have had to do this task manually by precisely anatomizing each hadith's chain of narrators or the list of people related to the transmission of a particular hadith. The evolution of modern computer science techniques has enabled new methods and introduced a potential paradigm shift in the science of hadith authentication. Focusing on the chain of narrators (also known as "Isnad") of a hadith, we have used a technique called `Sentiment Analysis' from Natural Language Processing (NLP) to build a text classifier which tries to predict the authenticity of a hadith. It learns from our custom-made dataset of Isnads and predicts an unknown hadith to be either authentic or fabricated based upon its Isnad. Our classifier was 86% accurate when tested on the test hadith dataset.