The way people learn has radically changed as a result of information technology. As an informal method of learning, fragmented learning has become a popular way to learn new technology and expertise. Academic organizations generate a large amount of heterogeneous data, and academic leaders want to make the most of it by analyzing the large amount of data in order to make better decisions. The volume isn't the only issue; the organization's data structure (structured, semi structured, and unstructured) adds to the complexity of academic work and decision-making on a daily basis. As big data has become more prevalent in educational settings, new data-driven techniques to enhance informed decision-making and efforts to improve educational efficacy have emerged. Traditional data sources and approaches were previously too expensive to obtain with digital traces of student behaviour, which offer more scalable and finer-grained comprehension and support of learning processes. This study provides a fragmented learning solution for students in a data environment that can suggest subjects to them based on their geographical location, gender, and district of residence, among other factors. This suggested framework is expected to play a key role in directing the development of a society that values lifelong learning.

Many recommendation systems make product or service suggestions based on existing knowledge of the user or the item. We must deal with two types of cold start problems: item-based and user-based cold start problems. In a user-based cold start dilemma, it is difficult for the system to suggest news to a new user whose information is not saved in the system. In this article, we attempt to address user based cold start problem by assuming that in the case of a user, we only know one type of information about the user, and that information is the user's location. Using BBC news data, an ID3 classification approach was developed, which incorporates eight explanatory factors such as News ID, News text, Keywords, date, Location, Shares count, followers, and so on. The classification accuracy of one of the best fit models constructed using (80-20)% training and test ratios is around 78%. Our technique is an effective tool for the cold-start problem because it outperforms the advice by a significant margin depending on the location. According to the results, our approach is competitive in terms of both accuracy and precision.