Mining Students' Data with Holland Model Using Neural Network and Logistic Regression

Education domain provides many interesting and challenging in data mining applications that potentially identified as a tool to help both educators and students, and improve the quality of education system. Nowadays, the impact of Minister of Education (MOE) regarding surplus graduates particularly from public universities somehow had an impact on Universiti Utara Malaysia's (UUM) undergraduate intake. As a result, students who applied to undertake a program at Faculty of Information Technology and Faculty of Management Technology come from various background. Hence this study aims to get some insight into first year students undertaking undergraduate program such as Bachelor of Information Technology (BIT), Bachelor of Multimedia (BMM) and Bachelor in Management of Technology (BMoT) at Universiti Utara Malaysia. The Holland Personality Model was used to indicate the students' personality traits. The study concluded that BIT students are not from the Social type since none of the Social personality type is significant. Most of BIT students have Arts background, expect a few who have sat for Perkom (Perkomputeran) subject during the STPM examination. As for the Holland Model, It also appears that BIT students are more Artistic since 50% of the questions that measure the personality type is significant. In addition, the BIT students are Realistic (33.33%) and Investigative (33.33%) type. The results also reveal that the BIT students concluded as Artistic, Investigative and Realistic (AIR) in personality types that ar ein accordance to Holland personality theory, this finding were also supported by Hansen and Campbell (1985) that suggested that Investigative, Realistic and Artistic (IRA) should be the code for computer professionals.