| Abstrakt: |
Today, measuring data quality is one of the most important strategies in improving data-driven business processes. Any correct decision to improve systems in this category of organizations depends on an appropriate analysis of data quality. Research data and especially theses/dissertations of graduates of the whole country based on this principle from various aspects of data quality need to be reviewed and evaluated. In the process of registering, the quality control of metadata is one of the most important parts that examines the information items of documents (such as the name of the researcher as well as supervisors and advisors, abstract, index, etc.). In the current situation, incompatible documents are identified during the quality control process and after entering in the system (as text), the relevant document is systematically returned to the researcher. Lack of a standard framework for identified non-conformities besides absence of an appropriate partitioning, make statistical analysis of metadata quality problems difficult, and also make it impossible to analyze the root of observed errors. Therefore, in this study, the incompatible structure observed after standardization has been used experimentally in two-month periods and the results have been presented. Incompatible in thesis defense date, the title page (Persian and English) and the existence of white pages in the file have been among the most important reasons for returning documents to users. Also, the analysis of all data showed that 59% of the incompatibles were related to the attached files and 41% to the information recorded in the system. Finally, based on the performed analysis, guidelines have been provided for the users of the system in order to improve the quality of data, according to specialized areas. [ABSTRACT FROM AUTHOR] |
