Data Preprocessing Techniques for Research Performance Analysis

Business intelligence (BI) system mixes operational data with the analytical tools to represent descriptive and complicated data to groups of decision makers. BI aims to enhance the features and accuracy of data warehouse to the decision-making process and widely applied in industry. In order to achieve that, BI pulls and gathers information from multiple sources of information systems. Data from multiple sources tend to have flaws such as missing values, inconsistency data, and redundant data. Hence, this paper aims to show data preprocessing techniques used to produce clean and quality data for Universiti Teknologi Malaysia (UTM) research performance analysis. For this research study, required data were provided by UTM management level. In future, this study is expected to compare different data preprocessing techniques and recommend the best one for research performance analysis.