Research Of Financial Firm Unstructured Information Retrieval System Based On Semantic

With the development of finance informatization level, more and more financial business are managed with the method of informatization. There are huge number of data unstructured information within finical firms. How to fetch valuable content from huge amount of unstructured information as soon as possible is the problem that we should face in the financial domain of Information Management. Although the Traditional Full-Text Retrieval Technique can quick match the information form keywords, but has some defects as follows: It can't integrate unstructured information which is in different data source and type; It can't analysis and reason the information that the user inquires; The search results also contain many valueless information and can't be digged with requirement.To solve these problems, We provide Unstructured Information Retrieval Method Based on Semantic. It is on the base of UIMA (Unstructured Information Management Architecture) Specification and Traditional Full-Text Retrieval Technique. At first, the method integrate different data source and data type into CMS (Content Management System) as a unified data platform. Then using extended UIMA Framework to analysis financial unstructured information resources and using Lucene to index and store unstructured information. On the base of traditional information retrieval model, we import ontology ideas and introduce the retrieval model based on ontological domain. According to create the OWL (Web Ontology Language) Standard financial domain's ontology database, we can realize information retrieve based on semantic. On the base of Unstructured Information Retrieval Method Based on Semantic, the strategy of Unstructured Information Retrieval System Based on Semantic are provided. According to the strategy, we design and realize the application system called"FUIRS"(Financial Unstructured Information Retrieval System). It consists of four parts which are unstructured information content management sub system, analysis sub system, content index sub system and associated search sub system. Unstructured information content management sub system is used to integrated and manage different kinds of unstructured information resource. Analysis sub system is used to get and analysis the data from content management sub system. Content index sub system can index and storage the data. Associated search sub system is used to retrieve information and provide the result to interaction platform for users.FUIRS's core module and capability are validated with unit testing and stress testing. Furthermore we validate FUIRS's function in the testing example. The results of the testing show that the strategy of Unstructured Information Retrieval System Based on Semantic is valid in practise.Compared with traditional Full-Text Retrieval System, FUIRS'features as follow: Firstly, it can be used to integrate different kinds of unstructured information source. Secondly, it based on financial firms build and support for the extension of business data analysis and application. Thirdly, using the ontology technologies based on OWL Standard, support semantic analysis and reasoning which made users get more complete and accurate information.