Dissertation > Excellent graduate degree dissertation topics show

Research of Financial Firm Unstructured Information Retrieval System Based on Semantic

Author: ChenBin
Tutor: RaoRuoZuo；CaoShunLiang
School: Shanghai Jiaotong University
Course: Software Engineering
Keywords: Unstructured information Ontology UIMA Full Text Search Semantic retrieval Financial enterprises
CLC: TP391.3
Type: Master's thesis
Year: 2011
Downloads: 72
Quote: 0
Read: Download Dissertation

Abstract

As the financial industry itself constantly improve the level of information technology, more and more financial services to information technology allows us to provide supervision and services. In these operations the existence of large amounts of unstructured data information, and how quickly the information from the mass of unstructured data to obtain valuable content and applied in the financial enterprises in the information management problems. Although conventional full-text retrieval technology to meet the information based on keywords to quickly find matching needs, but has the following disadvantages: Can not complete with heterologous heterogeneous data integration features of unstructured information; Unable query requests information on demand semantic analysis and reasoning; in the search results there are too many worthless and irrelevant information. To solve these problems, this paper Unstructured Information Management Architecture UIMA (Unstructured Information Management Architecture) specification and full-text retrieval technology basis, we propose a semantic-based unstructured information retrieval method. This method first of financial enterprises heterologous heterogeneous data sources to integrate unstructured information, and through the content management system CMS (Content Management System) for unified management of information resources. Then use UIMA framework to achieve scalable unstructured financial information for these resources data acquisition and expansion of data analysis, and application of technology to achieve the Lucene index content and results of the analysis of the data serialization index. In information retrieval, the method in the traditional search model, based on the introduction of the concept of ontology, we propose a retrieval model based on domain ontology, by building on OWL (Web Ontology Language) standard financial sector ontology-based semantic information Search. Based on semantic retrieval method based on unstructured information, this paper proposes a semantic-based financial enterprise unstructured information retrieval system design, according to the program design and implementation of an application system FUIRS (Financial Unstructured Information Retrieval System). FUIRS unstructured information from the content management subsystem, the analysis subsystem, content indexing subsystem, the associative retrieval subsystem is composed of four parts. Content Management subsystem is responsible for financial enterprises heterologous heterogeneous data integration and management. Analysis subsystem is responsible for the content management subsystem data and achieve scalable data analysis. Content indexing subsystem is responsible for indexing and storage of the data. Associative retrieval subsystem is responsible for semantic-based information retrieval, and provide users access to interactive search platform. By FUIRS systems, financial companies can effectively integrate unstructured information resources to achieve business data content analysis applications, and through ontology technology for the financial enterprise users with efficient data retrieval services. In this paper, unit testing, performance testing, two methods FUIRS core functionality modules and system performance is tested at the same time in the application according to its characteristics in the case of information retrieval function FUIRS data validation results show that semantic-based financial institutions Unstructured information retrieval system design is feasible in practice and effectively. And compared to the traditional text retrieval system, FUIRS system has the following features: the realization of effective integration of unstructured data sources and content acquisition; based on characteristics of financial enterprises build, support the expansion of the business data content of the data analysis and application; based standard OWL ontology technology, to support the retrieval of semantic analysis and reasoning capabilities, enabling users to obtain a more comprehensive and accurate information.

Related Dissertations

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Retrieval machine
© 2012 www.DissertationTopic.Net Mobile