Multimedia Answering and Retrieval System based on CQA with Media Query Generation

Abstract

The question answering system which has recently received an attention from the various information retrieval systems, machine learning, information extraction and the natural language processing the goal of the QAS is to retrieve the answer to the question than full documents. This question answering system which works on the various modules related only to the question processing, the document processing, and the answer processing. This QAS which doesn’t work properly with the main module which is questioning processing this system fails to categorize properly the questions. So to overcome the QAS the Community question answering (CQA) has gained popularity. As compare to QAS and automated QA sites the CQA sites are more effective. In this drawback available for community question answering system is that it only provides the textual answer. Here in this paper, we propose a scheme that enhances the textual answer with the multimedia data. The outline of Community question answering which mainly consists of three components: the selection of answer medium, the query generation for multimedia search and the selection and presentation of multimedia data. This approach automatically defines which type of media information should be added for the textual answer. Then it automatically collects the data from the web to supplement the answer.by handling an available dataset of QA pairs and adding them to a pool, in this, our approach is to allow a new multimedia question answering (MMQA) approach so as the users can find the answer in multimedia matching the questions pair those in the pool. Therefore, the users can approach MMQA from Web information will answer the questions in different media formats (text, video, and image) as particularly selected by the users.