| Summary: | Data Mining (DM) is an approach to discover useful information from large amount
of data. In education, DM plays an important role to discover hidden information
about students' learning behaviour and their academic performance. At the beginning
of a new semester, lecturers face difficulties because there is no approach that can be
used to identify the students' performance towards their studies. Appropriate actions
should be taken as early as possible to prevent the students from gaining low Grade
Point Average (GPA) at the end of the semester. This research aims to classifying
Students' Academic Performance (SAP) by applying DM techniques namely Naive
Bayes Classifier (NBC), Rule-Based (RB), and Decision Tree (DT). The dataset used
for the study is based on the students' particulars in the Bachelor programs in the
Faculty of Informatics and Computing, Universiti Sultan Zainal Abidin (UniSZA).
The data is of eight-year period beginning from July 2006/2007 to July 201312014. A
total of 399 records with eight parameters have been collected from the Division of
Academic Management, UniSZA. A Statistical Package for the Social Sciences
(SPSS) tool is used to identify the significant parameters that contribute to SAP and
Waikato Environment for Knowledge Analysis (WEKA) tool is used to conduct the
DM analysis. The data from the selected parameters is then cleaned, trained, and
tested to produce the best model to classify SAP. The extracted model is able to
forecast future academic performance of the first semester for the first year students
based on three GPA categories which are 'Good', 'Average', and 'Poor'. It is
discovered that, RB model shows the best result compared to NBC and DT models.
RB model produces the accuracy value of71.3% based on three significant parameters
which are students' race, gender, and family income. In conclusion, it is very useful
for the lecturers and the faculty management to embrace SAP model to be used in
identifying students' performance at the very early stage. A set of rules generated by
the model can also be used to classify the SAP for the new intakes. Hence, this
information provides assistance in terms of the necessary actions that can be 'taken to
improve the students' performance so as to guarantee success in the students'
academic tenure.
|
|---|