Evaluating Reasoning in Factoid based Question Answering System by Using Machine Learning Approach

in this paper; we present a closed domain factoid question answering system focusing on a why type Questions that can generate the answers of subjective type questions. The reasons we work on this research is that we want to show the system on what bases the why type Answers are generated this is because the strong requirement to satisfied in the early phase of system development. we follow the knowledge-based approach using TF-IDF algorithm and TF-IDF Vectorizer module we find that using the four parameter on the bases of because, neither, either, thoughts, and thus parameter is considered and lastly we present an evaluation of our system by giving precision, recall, & Accuracy of our system for why type questions.