Design And Implementation Of The List Of Classes Of Automatic Question Answering System

With the explosion of information, it has become a very important field of recent research to find an exact answer to a question in the large amount of information. Question answering system is such a system that receives the natural language question as input and returns the exact answer for the question. As there are usually more than one correct answer to a given question, this paper focus how to find as many correct answers as possible from an unstructured text corpus collection.This paper discusses the differences between list question answering and factoid question answering. According to these characteristics, we design a list question answering framework and implement a list question answering system. The system utilizes a keyword based classification method of answer type, a phrase based document retrieval model and an answer distance based ranking model in the framework.All the models and methods solve the most difficult problems in list question answering to a certain degree. Our system outperforms the baseline system significantly in the experiment.