Comparative Study of Multimedia Question Answering System Models

Abstract

Question Answering Systems is a mechanized way to deal with finding right solution to the inquiries posed by human in their own language. The enormous measure of data returned by web crawlers like Google, Yahoo, and so on, clients are over-burden to track down the right data from the web search tools. This QA approach takes care of these issues. Rather than returning a rundown of reports and connections from the momentum web crawler, the QA framework gives data from a bunch of replied, alongside reasonable media information. Question answering procedures plans to exploit top to bottom etymological and media content investigation just as area information to return accurate responses to regular language questions. QA framework has rundown of boundary like: question size, m-dimension, technique count, answer cardinality, and answer range. This paper investigates the advancement of mixed media question replying research, their plan boundary, and subtleties its future perspectives. Multiple Line Questions and Video Question Answering methods have been discussed at length and their performance measured on QA framework parameters.