SERVICE COMPUTING SYSTEM ENGINEERING FOR STATISTICAL METADATA BASED ON SERVICE ORIENTED ARCHITECTURE (SOA)

• Main Author: RAHIM, RAHMIATI
• Format: Theses
• Language: Indonesia
• Subjects: Teknik (Rekayasa, enjinering dan kegiatan berkaitan)
• Online Access: https://digilib.itb.ac.id/gdl/view/30153
• Institution: Institut Teknologi Bandung

<p align="justify">Statistics Indonesia – BPS is a non-ministry institution that is assigned as a Statistical Activity Organizer’s Coordinator in Indonesia. Based on the role, BPS tries to integrate the all of the statistical activity such as census or survey from silo based system to be integrated by constructing corporate statistical infrastructure based on international standard. One of the goals tried to be achieved is the the managed metadata governance. At the moment BPS has not adopted any proper metadata management principle. The problem is metadata have not been maximized to it’s potential to support statistical business process such as planning, designing, building, up to the stage of evaluation. It causes statistical activity conducted is prone to be duplicate and inaccurate. Metadata are created and used in each fase of statistical business process so that a proper technology support is needed to assure that metadata could be stored and accessed from a single source. Service oriented model approach like Service Oriented Architecture (SOA) is adopted to support interoperability across institution and platform. <br /> <br /> The research design utilizes Service Computing System Engineering Framework as the methodology. The framework is valued to be able to accommodate a comprehensive engineering steps from requirement analysis to development. The research also has developed engineering activity in each phase of the framework up to the stage of Development. <br /> <br /> The objectives of the research are to design and develop a metadata statistical service computing system based on SOA. Statistical metadata service computing system is a business metadata repository regarding statistical activity which integrates all of metadata object in any domain and adopts management principles that complies to standard. Statistical metadata service computing system would serve the necessity of metadata objects in each of statistical activity phase. <br /> <br /> The results of the research are the design of metadata management business service which integrated with BPS’ statistical business process, data architecture design, and integrated IT services architecture. Service computing system design is validated based on measurement metric of SOA design principles. It resulted that the proposed design has low value of coupling factor, high cohesion metric, high value of reusability, and low complexity value. Prototypes are developed based on validated design in which further tested with unit testing, composite testing, and system testing. A testing also conducted to the end user by using User Acceptance Testing (UAT). It is resulted that the proposed system has conformed enough the business service requirements.<p align="justify">