MULTIMEDIA QA GENERATION BY USING SEARCH DIVERSIFICATION?

Community Question answering (cQA) services have gained popularity for the past few years. It supports community users to post and answer questions and also it enables general users to acquire information from a set of answered questions. Though existing cQA forum provides textual answers alone, it is not much informative for many questions. In order to enhance textual answers in cQA with suitable media data MMQA method has been introduced. This method consists of three parts, Answer medium selection, Query generation for multimedia search and multimedia data selection and presentation. This method automatically finds out which type of multimedia information should be added to get an elaborated textual answer and also it automatically gathers data from the web to enhance the answer. By processing a collection of question answer pairs and adding them to a dataset it can set up a novel multimedia question answering (MMQA) method as users can find multimedia answers by comparing questions with those in the dataset. This MMQA method not only provides image and video for direct question answers but also give answers for more complex questions. The multimedia search diversification method is used here to collect the relevant answers based on questions. The result shows that it provides more satisfactory answers to the users and also it is more effective.