Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2022
- Goals
- Record
- 1781ce15-6d1a-4dae-bb38-08a671034172
- Published
- 2026-10-30
Universitas Dian Nuswantoro (UDINUS) conducted a study on the development of a big data ecosystem using multiple sources of higher education data in Indonesia. The activity involved UDINUS researchers Yunita, Ariana; Santoso, Harry B.; and Hasibuan, Zainal A. The study addressed the challenge of anticipating appropriate action for student achievement, particularly because students may have different levels of understanding and because interventions may occur too late when a student is experiencing difficulty.
The activity gathered possible features from multiple data sources, including national education databases, reports, webpages, and other sources. The dataset comprised information on undergraduate students from 13 provinces in Indonesia. The student-related information included academic histories, demographic profiles, and socioeconomic backgrounds. The activity also included institutional information concerning accreditation level, programmes of study, type of university, and geographical location.
The gathered data was preprocessed through techniques addressing missing values, data categorisation, data consistency, and data quality assurance. This process produced a relatively clean and sound big dataset. Principal component analysis was then employed to reduce the dimensions of the dataset. K-Means methods were subsequently used to identify clusters or inherent structures within the dataset.
The K-Means analysis suggested seven clusters: very low-risk students, low-risk students, moderate-risk students, fluctuating-risk students, high-risk students, very high-risk students, and fail students. The study also revealed gaps between public and private universities across three regions in Indonesia, between STEM and non-STEM programmes, between rural and urban contexts, and according to accreditation status. A gap in the distribution of quality human resources was also identified.
The activity supports SDG 4 by examining student profiles, educational contexts, and student achievement through a structured analysis of higher education data. Further study was identified as necessary to use the characteristics of each cluster to predict student achievement and provide solutions and intervention strategies to improve likely student success. Through this activity, UDINUS frames the use of multiple data sources and analytical methods as an approach for understanding differences in higher education student circumstances.