Design And Implementation Of Structured Query Subsystem Of Financial Questions System

In recent years, search engines have become the major field that many companies involved in, the search is divided into two areas, one area is the whole field of the search, which almost covers people’s basic life. Another area is the professional field search, which mainly specialize in one area to explore. Currently, the number of the search engine for the financial sector is very few, in order to allow users to search the information that they want to know about the stock in accordance with their wishes, SouNiu project team raise IWenCai project, which is committed to provide users with more efficient, more professional, more accurate financial quiz experience.This article focused on the financial sector, which about stock selection for natural language processing research problem, finally finish the development and the test of the system. The system contains the main questions to analysis and search. Analysis of some of the major questions are based on information that has been generated query semantic tree into the information for the search, finally send the message to the users. The system developed using the Java language, specific modules include the following six sections: parse structured list of conditions, thematic analysis, indicators accumulate, data query, anti-analytical and analytical Web client Web interaction data. Through this six-part design, the system completed a structured query to return data from a semantic tree.Currently involved in the search database including relational databases and non-relational databases, when faced with a large amount of data search, relational database search’s efficiency is very low, rather than non-relational database have their own feature. In this paper, it processes the semantics tree of natural language processing to the structured storage. Through the use of different data sources for different types of data, it makes semantic tree easier in the application and more efficient to store and read. Because different application types for each non-relational database approach in the implementation process calls for a variety of non-relational database package, which ensures that any changes in the subsequent maintenance of the database will not affect the structural condition of the tree section, and a variety of database calls package also improve code reusability. This project uses a lot of optimization and design patterns, thereby improving the efficiency of the system and post-optimization.Finally, all of the function and performance pass the test to ensure the integrity and performance of the system, the system has achieved the desired goal. At present, the system has been on-line. According to the statistics of the user query, the count is about50000after getting rid of repeat, the proportion of parse correctly is about92percent. It has got the expected requirement.