Design And Implementation Of Chinese Automatic Question Answering System Based On Search Engine

How to quickly and accurately obtain the necessary information from the mass of information has become an important issue. The search engine is popular because of its convenient and quick response characteristics, and become the most important way of information acquisition. However, it is difficult to clearly express the intention of the user that search engines retrieve information with key words. On the other hand, it still needs to manually find the answer from the related web pages. So, the automatic question answering system emerge as the times require, the traditional automatic question answering system not only has the information which is small in coverage and slow in update, but also needs to maintain a large database. In order to play the strengths of both, this paper intends to combine the two together to design and realize a Chinese automatic question answering system based on search engine.The main work of this dissertation is as follows:(1) Improved Site Q, and proposed Topic-Site Q. For their greater contribution to the semantics of the paragraph, this paper puts the first and the last paragraph and sentences into Site Q with appropriate weight, and proposes Topic-Site Q: using a multifeature fusion algorithm to compute the semantic similarity of the first and the last sentences that contribute to the paragraph similarity by weight. At the same time, we improve the score of the first and the last paragraphs, recommend and sort the paragraphs by the final score and return the candidate paragraph.(2) Improved an answer extraction algorithm based on semantic dependency. The answer extraction algorithm based on semantic dependency mainly studies the semantic and syntactic structure, and the inspection method is single. Word frequency is one of the important characteristics, and should be reflected in the answer extraction algorithm. We improved the answer extraction algorithm based on semantic dependency, considered the word frequency into it, and use a log linear model to fuse the two together.(3) Designed and implemented a Chinese automatic question answering system based on search engine, and according to the two optimization methods presented to optimize the system. First, we analyze the need of the system in detail. Second, describes the overall structure of the system and gives the system structure diagram. In the part of detail design and implementation, we discussed the functions, the processes, the implementation details and the main algorithm and its improvement of each module.(4) In order to verify the effectiveness of the proposed optimization method, this paper artificially construct the problem test set to test the improvement effect of the algorithm and the system, and then calculated and compared the precision, recall and F1 value of the two improved algorithms and the system using the improved algorithms.The experiments show that, the optimized algorithm has a good effect, after using the improved algorithm, the performance of the system is much better than before.