A Group-Based Feature Selection Approach to Improve Classification of Holy Quran Verses

Automating quranic verses labeling using machine learning approach

“…The input data are Quranic verses extracted from the combined sources of Holy Quran translation and tafsir. The study identified the significance of combining multiple related Quranic sources [1,11] for better understanding of the input verses.…”

“…The input verses are grouped into one of three predefined labels: faith, worship, and etiquettes. These class labels are from the most fundamental aspects of Islam [1,3].…”

“…As the world advances over the years with massive technological growth, it becomes more demanding and necessary for AI systems to be able to make decisions automatically and independently within a time frame. In order to achieve this, computers need to learn without being explicitly programmed [1].…”

“…The Quranic text is an important holy book of Muslims' faithful [1,[9][10]. It is one of the most widely read and referenced resource.…”

“…Research in automating the Quranic text has gained attention in recent times. Some of the existing works as found in literatures include: text classification applications on the Holy Quran [1,3,[11][12][13]; ontology-based applications [14][15][16][17]; digitized Holy Quran applications [18][19][20][21][22]. Furthermore, conventional among machine learning algorithms often implemented in ML tasks include: naïve bayes (NB) [4], decision trees (J48) [23], neural networks [24], support vector machines (SVM) [25], and k-nearest neighbour (k-NN) [26].…”