Revolutionize Cosine Answer Matching Technique for Question Answering System

Imitating humans by computers i.e. Knowledge Engineering has been into trend from past few years, resulting into reducing the unnecessary manpower and skilled workers in the field of education when it comes to computer assisted assessments. Various techniques developed to evaluate answers by computers are known to be non-performing resulting into undeveloped system that revolutionizes traditional method of assessment and instant evaluation of results. One such technique is the TF-IDF using cosine similarity which proves to be a bit lacked out when comes to synonym correction. This paper aims at revolutionizing the traditional assessment technique of pen paper, replacing it with modern solutions where computer imitates the skilled person. The main objective is to achieve an algorithm that best represents the or imitates the human when it comes to evaluating the subjective answers using Natural Language Processing techniques. This method proves to be efficient enough to withdraw the traditional system and develop a distinct solution that will benefit various Educational Institutions.