Intelligent answering location questions from the web using molecular alignment

摘要

In this paper, a new molecular alignment based recognition method for question answering from from the Web is proposed. This identifies locations using an molecular alignment sequence algorithm according to their similarity with a user natural-language question. Different experiments and results concerning questions on locations are discussed. The high accuracy of the proposed alignment strategy shows the promise of approach to effectively deal with questions extracted from natural-language corpus which contain many complex patterns.