Study Of Collaborative System Integration For Multi-Source Heterogeneous Data In Multiple Formats Enterprises

In the age of the Internet, it is the inevitable trend that using the means of information technology to management enterprise. The data acquisition is the basis to realize the enterprise information management. But with the rapid development of the enterprise, formats in enterprises are increasingly rich; the heterogeneous systems in the enterprise are become much more, the data presents characteristics like multi-source and heterogeneous and so on. This phenomenon is particularly significant in the group enterprises. To realize the data collaborative system integration, is to acquire data from multiple formats group enterprise multi-source heterogeneous systems automatically, completely and originally,avoid the "decoration" and modifies of data, do effectively knowledge discovery of data and provide support for the daily management and management decisions of enterprises. To achieve these goals, realize he data collaborative system integration effectively, there are four types of common problems to be solved:The first one is heterogeneity problem, including heterogeneous hardware environment, heterogeneous operating systems, heterogeneous file systems, heterogeneous database and heterogeneous data structure, etc. The second one is the on-demand problem. Data integration needs to access the data as much as possible on demand. The higher the on-demand, the level of integration is higher, the function of the system is more perfect, and the access to data is more convenient. The third one is data security problem. In the integration systems, how to ensure the accuracy, stability, and high availability of data transmission, transformation, processing, etc, and avoid various system abnormal situations which result in data loss or inconsistent are difficult problems. At the same time, data access control is also a difficult problem. The last one is data presentation and portal optimization problems after integration. It needs effective knowledge discovery to integrate data, and to take users or employees’ use habits and related experience seriously.Aiming at heterogeneity, on-demand, data security, the optimization of the portal problems above in the data collaborative system integration, the dissertation based on the existing research results and works, for multi-source heterogeneous system in multiple formats enterprises, carries out the research surrounding the multi-source heterogeneous data sources collaborative system integration, does the following innovative work.1)The dissertation puts forward a kind of multi-source heterogeneous data integration middleware method (MSHDIM, Multi Source Heterogeneous Data Sources Integration Middleware),for the collaborative system integration problems involved the multi-source heterogeneous data integration. This middleware as the business layer is built between the data layer and application layer and provides the business logic services for data integration; every heterogeneous data source system constitute the data layer of system; the application programs in application layer constitute the presentation layer of system. In this way, the metadata level operation is avoided, it does not need to deal with complex mapping relation about local model and global model and deploy core Web service in the data source layer etc.,. The method uses the JDBC/JDBC-ODBC Bridge driven in MSHDIM, achieves the multi-source heterogeneous data source connection through the driven management. This method also involves dynamic pool connection model, avoids the integration effects on the performance of the original data source and realizes data source dynamic addition. This connection method does not require the global mapping conversion of original data and maintains the characteristics of the original data sources, also save the integration time conveniently.2) The dissertation puts forward data extraction conditions and process based on MSHDIM method for the collaborative integration on-demand data problems. The system automatically determines whether the extracted data in the virtual table namely view or not, if not, creates view and meets the user’s on-demand data requirements. In terms of data change, MSHDIM method also presents modify data conditions and process. In the mean time, MSHDIM method calls the Resin application service, generates dynamic content on collaborative system page through the use of JSP labeling or the small script. Contents are encapsulated in the labeling and JavaBeans components, also in the small script, all the scripts run on the server and grant mobile intelligent terminal authorized access. In this way, the collaborative integration on-demand data problems are solved.3) The dissertation puts forward a method which data sources generate different views automatically according to different access permissions, and then extracts the relevant data in MSHDIM method according to the access permissions, for the security problems of collaborative integration system data and mobile intelligent access terminal. This method solves the problem of accessing authorized data. As mentioned in the dissertation the data source of the system integration is the real-time online system, this method also avoids the risk of directly access into the database table. Because of to the character of the collaborative system integration, the mobile intelligent terminals need to access to collaborative integration data on-demand at anytime and anywhere, so this dissertation puts forward a security method to solve the intelligent mobile terminal data security and anti-theft problems. This method can send users’registered account, mobile intelligent terminal IMEI, the new SIM card IMSI and other information to collaborative system (data server) integration server, via collaborative system integration server push the APP uninstall instruction and delete related directories and files in the mobile terminal to achieve the effect of safety and security.4)The dissertation combined with the characteristics of collaborative system integration portal, does association rule discovery, puts forward a method to reduce the candidate sets, and optimizes the Apriori algorithm for the collaborative system integration portal optimization problems. The improved algorithm performance and stability are more optimal than the ordinary Apriori algorithm, and more suitable to collaborative system integration portal optimization. Through the relevant log from collaborative system integration, the dissertation mines knowledge of the collaborative portal, optimizes the collaborative system integration portal.