Multi-Label Classification System That Automatically Tags Users' Questions To Enhance User Experience

Abstract

Information tagging is critical to the indexing process. Web portals based on question response mechanisms include Stack Overflow. There's a lot of information here, and it's arranged using tags. The study's goal is to develop a system that relies on self-tagging to track objects. It makes advantage of the 'Document-Term Matrix' idea to foretell different tags connected to a problem. This is accomplished by selecting all tags with probabilities greater than a predetermined threshold. The research contributes to demonstrating how machine learning models may be put to use. By doing so, it also creates a statistical link between accuracy and the amount of inquiries per tag. There is a benefit in optimising the parameters, such as how many tags there are for each question.