New level of information sharing is enabled by different online communities like wikis, blogs, forums etc. which provides a platform for interaction with individuals which offers services like searching and posting queries or answers and share expertise with other information seekers. For recently posted query or searched query system furnishes the pool of answers with similar questions links, which could be a prolonged task for finding the significant answer. To overcome this, the system proposes an approach to effectively rank answers which are most relevant and best from historical archives based on similar queries found. It comprises of two components, one which contains training samples with positive, negative and neutral classes and other component retrieves similar questions to posted questions which are with their answer pools. Two data mining approaches were compared to retrieve similar questions. Our objective is to rank answer candidates based on pair wise comparison where question-Answer pairs are ranked using pair wise learning to a rank approach based on a trained model which provides the user with most relevant answers for a given posted question. © 2018 IEEE.