Abstract —Due to increase in the volume of students’ data and the limitations of the available data management tools, higher education institutions (HEIs) are experiencing information overload and constrained decision making process. To attend to this, Information Visualization (InfoVis) is suggested as a befitting tool. However, since InfoVis design must be premised on a pre-design stage that outlines the explicit knowledge to be discovered by the HEIs, so as to actualize a functional and befitting InfoVis framework, this study investigates the explicit knowledge through survey questionnaires administered to 32 HEI decision makers. The result shows that relationship between the students’ performance and their program of study is the most prioritized explicit knowledge, among others. Based on the findings, this study elicits a comprehensive data dimensions (attributes) expected of each data instance in the HEI students’ datasets to achieve an appropriate InfoVis framework that will support the discovery of the explicit knowledge. Our future work therefore include designing the appropriate visualization, interaction and visual data mining techniques that will support the explicit knowledge discovery and HEI students’ data-driven decision making types. Index Terms—Explicit knowledge, HEI students’ data-driven decision making types, InfoVis, knowledge discovery. I. INTRODUCTION The ever-growing nature of students’ data of HEIs has accounted for the experience of information overload, and subsequently its constraint in supporting decision making processes. This experience has restricted HEIs’ decision makers and administrators in their pursuit of making wealthy use of the students’ data. To attend to this, information visualization (InfoVis) is arguably the suitable tool that harnesses its strength in gaining insight from large multidimensional datasets and therefore aids decision making [1], [2]. Notably, InfoVis is a research domain on data analysis and knowledge discovery through visual exploration [3], [4] in view of supporting decision making. However, it is said that InfoVis cannot be functional without a pre-design stage that thoroughly investigates the explicit knowledge expected to be delivered [5], [6]. Identifying the explicit knowledge involved in the application areas of InfoVis will guide the transformation of the raw data (structured or unstructured) to meaningful information that will assist in decision making and creation of knowledge and wisdom [7]. Considering HEIs, the experience of information overload and the limitation in the currently used data management Manuscript received June 12, 2014; revised September 17, 2014. The authors are with School of Computing, Universiti Utara Malaysia, Sintok, Malaysia (e-mail: ayobami.sm@gmail.com, zulie@uum.edu.my). tools demand a specific InfoVis framework in the domain [8]. This must also be preceded by a study to identify the explicit knowledge expected to be utilized by its decision makers and administrators [5], [9]. This study is therefore aimed at identifying the explicit knowledge involved in HEIs students’ data-driven decision making processes. Section II and II of this paper discuss the motivation of this study and past related works on students’ data-centered knowledge discovery goals, respectively. Section IV discusses the methods used in discovering the explicit knowledge; Section V discusses the findings, while Section VI serves as the conclusion. II. MOTIVATION OF THE STUDY Having understood the diverse applicability and indispensable usage of InfoVis, this study is motivated by the lingering problem of information overload [10] in the course of managing the data of higher education institutions (HEI). The HEI’s data is becoming increasingly difficult to analyze in view of getting previously unknown information that could assist policy and decision making in the education sectors due to its fast growing trend [11]–[14]. The currently used data management systems by the HEI are the traditional management information systems which are faced with problems of money and time wastage, loss of scientific and industrial opportunities, inability to explore heterogeneous data sources and subsequent exploit of its hidden information [15], [16]. Added to this is the usage of static statistical graphics like pie chart, bar chart and line chart. However, the multidimensionality of these datasets, its heterogeneous sources and superabundant volume has limited the success of these traditional data management systems and data representation using the statistical graphics only [17]. Other recorded cases are HEIs occasional deployment of popular statistical tools like SAS and software systems like Tableau for their operational data analysis [13]. But, the one-cut-fit-all model type of the software tools pose ineffectiveness in the decision support functionalities of the data management tools. This is because the software model is designed to fit all organizations irrespective of the difference in their intending explicit knowledge, and consequent data models. Table I shows the limitations of past works on the management of the HEI’s data. Sarker et al. [18] emphasized the role of institutional repositories in addressing the data management challenges of higher education. Their study identified that the inability to adopt new technologies, improve the quality of learning and Students' Data-Driven Decision Making in HEI: The Explicit Knowledge Involved Semiu A. Akanmu and Zulikha Jamaludin International Journal of Information and Education Technology, Vol. 6, No. 1, January 2016 71 DOI: 10.7763/IJIET.2016.V6.661