An English Islamic Articles Dataset (EIAD) for developing an IslamBot Question Answering Chatbot

Abstract

A chatbot is one of the most vastly recommended technologies to be used during these decades, especially through the digitization era. It could save much consumed time for both the users and the customer service employees. Chatbots could provide an answer to the asked questions instantly. IslamBot is an Islamic religion chatbot “i.e.,” responsible for answering any inquiries related to the Islamic religion. The aimed audience is non-Muslims people willing to join Islam or New-Muslims. Building such types of chatbots need to have an enormous amount of trusted data. Accordingly, in this paper The English Islamic Articles dataset (EIAD) is proposed as a benchmark reference for English Islamic question answering. So, this dataset contains about 10000 English Islamic articles. It is scrapped from authenticated and trusted websites like NewMuslims.com [1] IslamReligion.com [2], and IslamQA.com [3]. The dataset is about 275 articles from NewMuslims.com [1], 1550 articles from IslamReligion.com [2], and 8292 articles from IslamQA.com [3]. The EIAD dataset is a structured dataset “i.e.,” labeled and categorized. This dataset contains about 15 different categories. Each category is covering several different topics. This paper focuses on discussing how The English Islamic Articles dataset (EIAD) has been collected.