IMPROVING DATA QUALITY IN OPEN GOVERNMENT DATA MANAGEMENT PLATFORMS THROUGH THE IMPLEMENTATION OF DATA GOVERNANCE (CASE STUDY: DIRECTORATE GENERAL OF FISCAL BALANCE)

• Main Author: Habibie, Khairul
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/70441
• Institution: Institut Teknologi Bandung

Open government data is currently one of the government initiatives related to realizing data to create good governance. In financial administration, the Directorate General of Fiscal Balance (DJPK) has made data governance as outlined in the DJPK regulations to support the implementation of One Data Indonesia as a form of open government data policy in Indonesia. To support the implementation of data governance, a platform is needed to support any changes that may occur in data governance. Currently, DJPK has used CKAN as a platform to publish data but has yet to, as a whole, can support data management activities contained in DJPK data governance. In addition, after evaluating the published data, it turns out that there are data quality problems. For this reason, it is necessary to develop several data management activities in CKAN to support the implementation of data governance and improve the quality of published data. If the data meets data quality, the platform can ensure that the data to be published meets the quality of the data domain so that if there is data that does not meet it (according to the metrics set), data quality will be improved. The development of an open government data management platform in this study uses the design science research methodology (DSRM) methodology adopting the principles of Service Oriented Architecture (SOA). Based on the results of the design evaluation that has been carried out, the resulting design is by SOA principles, namely: coupling, cohesion, complexity, and usability, the design evaluation carried out and the results obtained are 0.0046, 0.9218, 0.0050, respectively and 2.5714. From this score, it can be interpreted that the design of the government data management platform adheres to SOA principles: loose coupling and high cohesion. Data quality testing is done by calculating the Valid DQI (positive value) and Invalid DQI (negative value) of the 7-dimensional quality data. The test results show an increase in the quality of the uniqueness, accuracy, integrity, conformity, and validity of data on the realization of the monthly APBD of XYZ District. With each result, there is an increase in the uniqueness of the valid DQI from 98.7203 to 100. The quality of conformity data has also increased from the previous Valid DQI value (positive value) of 94.7368 to 100. Furthermore, the dimensions of data quality accuracy have also increased significantly. Previously the Valid DQI (positive value) value was 0 to 100 by manually checking from custodians by uploading the pdf document used as a comparison. As for the dimension of data quality integrity, there was an increase from the previous Valid DQI (positive value) value of 66.6667 to 100 obtained from the data quality dimension values of completeness, accuracy and consistency, respectively, namely 100, 100 and 100, the integrity dimension value was 100. As a concern, the quality of the validity data is also different from the stage before the elimination of duplication. The percentage of the previous expenditure budget was 3.511849 to 2.04304 and was by the results of manual checking. On other data quality dimensions, completeness is worth 100, indicating no empty columns; consistency, which also has a 100, indicates that the data is aligned after being compared with existing reference data.