| The current on question answering usually achieve the answer only from unstructured text resources such as a collection of news or pages. According to our observation to real questions from yahoo! Aswer, users sometimes ask complex natural language questions which contain structured and instructured features. Generally, answering the complex questions needs to consider not only unstructured resource but also structured resource. In this work, we propose a new idea to improve accuracy of the answer of complex qustion by recognizing the structured and unstructured features of question and integrate structured and unstructured resources in the web. Our framework consists of three parts: question analysis, resource Discovery and analysis of the Relevanant Answer. In Question Analysis. Resource Discovery and analysis of the Relevant Answer In Quastion Analysis we used a few assumptions and tried to find structured and unstructured features of the question. Structured feature tend to structured data and unstructured feature tend to unstructured data. In the resource discovery we integrated structured data (relational database) and unstructured data (web page) to take the advantage of two kinds of data to improve and to reach the correct answers. We can find the best top fragments from context of the relevant web pages in the relevant Answer part and then we made a score matching between the result from structured data and unstructured data, Then finally we used QA template to reformulate the questions. |