DEVELOPMENT OF A BEST ANSWER RECOMMENDATION MODEL IN A COMMUNITY QUESTION ANSWERING SYSTEM

Abstract

Question and Answering (QA) systems have been the major tool being used by community users to get quick response to challenges. A typical QA system houses thousands of questions and answers from different scholars of diverse field of studies providing solutions to problems. A Community Question Answering System was developed to implement the best answer recommendation model proposed in this research work. The system applies Brouwer Fixed Point Theorem to prove the existence of the desired voter scoring function and Normalized Google Distance (NGD) to show closeness between words before an answer is suggested to users. For each question, answers are ranked according to their Fixed Point Score (FPS) and the highest scored answer is pronounced as the FPS Best Answer (𝐵𝐴). For every question asked by user, the system will activate NGD to check if similar or related questions with best answer had been asked and stored into the database. Brouwer Fixed point is then used to calculate the best answer from the pool of answers on question and the best answer is stored in the NGD data-table for recommendation purpose. The system was implemented using PHP scripting language, MySQL for database management, JQuery, and Apache. The performance of best answer recommendation system was evaluated using some of the standard metrics, which are: Reciprocal Rank, Mean Reciprocal Rank (MRR) and Discounted cumulative gain (DCG). The system helps to eliminate longer waiting time to receive answer in a community question and answering system.