1.1 Introduction
Student academic performance prediction is a critical aspect of educational data mining (EDM), which focuses on applying machine learning techniques to analyze and forecast students' future academic outcomes. According to Romero and Ventura (2020), educational data mining involves the use of computational methods to discover patterns in educational data that can help institutions make informed decisions. Predicting students' academic performance using decision tree algorithms allows educators to identify students at risk of poor performance and implement early interventions (Han et al., 2019). Studies has shown that the ability to predict students' performance is essential for enhancing teaching strategies, curriculum development, and personalized learning approaches. Machine learning techniques, such as decision trees, offer a systematic and efficient way to analyze historical academic data and predict future outcomes. Decision tree algorithms classify students based on key attributes like attendance, assignment scores, and previous academic records, providing a clear and interpretable model for prediction (Patil & Sherekar, 2018).
As a prelude to other parts of this study, this chapter will discuss the background upon which this study was initiated, the statement of problems that led to this study, the Aim and Objectives of the study. Others are Significance of the study, Scope of work, Limitation of the study and Definition of technical terms.
1.2 Background of Study
Since one of the goals oftertiary institutions is to contribute to the improvement of the quality and standard of higher education, the success in the creation of human capital has been a subject of continuous analysis. Hence the prediction of students' success is very important to these higher education institutions, because the purpose of any teaching process is to meet students' educational needs and enhance overall student's academic success. In this regard, important data and information are gathered on a regular basis after which they are used in the prediction of students' academic performance (Edin Osmanbegovic, 2012). Measuring and predicting the academic performance of students has been a challenging task since students' academic performance depends on diverse factors such as personal, socio-economic, psychological and other environmental variables. But the prediction of student's performance is a very important endeavor as it helps the student and teachers to minimize poor academic performances and produce better educated and enlightened students in order to make the society a better place. With the help of performance prediction, a failing student can be identified and helped by putting all the factors affecting the student into consideration and providing solutions to counter this factors so as to facilitate better performance (Brijesh Kumar Bhardwaj and Saurabh Pal, 2011).
According to Romero and Ventura (2020), EDM enables institutions to extract useful insights from academic records, helping educators and policymakers develop strategies to improve learning outcomes. Among various machine learning techniques, decision tree algorithms have emerged as a popular tool for classifying and predicting student performance due to their interpretability and efficiency (Han et al., 2019). The prediction of student academic performance is essential for identifying students at risk of failure, improving teaching methodologies, and enhancing personalized learning experiences. Educational institutions collect large volumes of data related to students' attendance, assignment scores, participation, and exam performance. By utilizing decision tree models, these data points can be systematically analyzed to uncover hidden patterns and predict future academic outcomes. Studies have shown that decision tree-based models provide accurate and transparent classification of students based on key attributes such as study habits, socio-economic background, and prior academic achievements (Patil & Sherekar, 2018).
Quinlan (2014) stated that, decision tree is a widely used classification algorithm in machine learning due to its simplicity and interpretability. It creates a tree-like structure where each node represents a decision based on a specific attribute, leading to a predicted outcome. In the context of student performance prediction, decision trees help in categorizing students into different performance levels based on their academic data. Studies have shown that decision trees can achieve high accuracy in predicting students' performance when trained with relevant datasets (Quinlan, 2014). Decision trees work by breaking down complex decision-making processes into simple, tree-like structures where each node represents a decision criterion based on an attribute. This structured approach makes it easier to interpret and apply predictions in educational settings. As Quinlan (2014) points out, decision tree models like C4.5 and ID3 offer robust classification techniques that can be adapted to various academic datasets to improve predictive accuracy.
The challenges encountered that led to the execution of the research work is that, many predictive models use complex machine learning techniques that may deliver high accuracy but lack transparency in decision-making. Decision trees, on the other hand, offer a more interpretable approach, making it easier for educators to understand why a student is classified as at risk and take appropriate action (Quinlan, 2014). It is against the background that the developments of this software will help educational institutions identify at-risk students early and implement timely interventions.
1.3 Statement of Problems
Investigation revealed that predicting academic performance is the complexity of student data, which includes multiple factors such as attendance, participation, prior grades, socio-economic background, and learning behaviors. Manually analyzing these factors to determine patterns and predict future performance is time-consuming and prone to human bias. Decision tree algorithms offer a structured way to process these datasets, but many educational institutions lack the necessary tools to implement them effectively (Han et al., 2019).
Furthermore, the lack of data-driven decision-making in education also contributes to inconsistent academic interventions. Without an effective prediction system, institutions rely on generalized solutions rather than personalized strategies to support struggling students. Studies by Patil and Sherekar (2018) show that using decision tree-based predictive models allows educators to tailor interventions based on individual student needs, improving overall academic success rates.
Lastly, without adequate measures to curb the existing problem of persistent students' failure, it will continue to remain a major problem for higher institutions. But with the analysis of the factors which are socio-economic, psychological and environmental, headway can be made towards curbing the problem of student failure.
1.4 Aim and Objectives of Study
The aim of this study is to design and implement a student academic performance prediction system using a decision tree algorithm to enhance early identification of at-risk students and support data-driven educational interventions.
The specific objectives of the study include:
- To develop a decision tree-based model that analyzes student performance data and classifies students based on their academic risk levels.
- To identify key factors influencing student academic performance, such as attendance, participation, and prior grades.
- To evaluate the accuracy and effectiveness of the decision tree model in predicting student performance.
- To provide an interpretable predictive system that allows educators to make informed decisions about student support and intervention strategies.
- To improve academic performance monitoring by integrating predictive analytics into the educational decision-making process.
1.5 Significance of Study
The implementation of the proposed system will provide an effective approach to predicting student academic performance using a decision tree algorithm, which will help educational institutions identify at-risk students early and implement timely interventions. Also, the research study will benefit educators by enabling them to analyze key factors influencing student performance, such as attendance, participation, and prior grades.
Furthermore, students will benefit from the study, as early identification of academic risks will lead to targeted interventions that will enhance their learning experience and performance. The implementation of this predictive system will foster a proactive academic environment where potential challenges will be addressed before they impact students' success.
Lastly, this research will contribute to the growing field of educational data mining and machine learning applications in education. It will provide a foundation for future studies that aim to refine predictive models for academic performance, ensuring continuous improvements in educational monitoring and student support systems.
1.6 Scope of the Study
This study will focus on the design and implementation of a student academic performance prediction system using a decision tree algorithm at Lagos State University (LASU), Lagos, Nigeria. The research will involve collecting and analyzing student performance data, including attendance, participation, previous grades, and other academic factors, to develop a predictive model that classifies students based on their academic risk levels.
The study will be limited to undergraduate students within selected faculties at LASU, ensuring that the dataset used is relevant to the university's academic structure. The decision tree model will be evaluated for its accuracy, effectiveness, and interpretability in predicting student performance.
1.7 Limitations of the Study
During the course of this study, many things militated against its completion, some of which are:
- Time Constraint: The time frame given to accomplish this project was very short due to school academic calendar and it was carried out under pressure which made the researcher not to implement some necessary features.
- Financial Constraint: Insufficient fund tends to impede the efficiency of the researcher in sourcing for the relevant materials, literature or information and in the process of data collection (internet, questionnaire and interview).
1.8 Definition of Terms
Academic Performance:
Academic performance refers to the level of achievement students attain in their educational pursuits, often measured through grades, test scores, and other assessment criteria (Hattie, 2009). It reflects a student's ability to grasp and apply knowledge in various subjects.
Prediction Model:
A prediction model is a statistical or machine learning-based approach used to analyze historical data and make informed forecasts about future outcomes (Bishop, 2006). In this study, the prediction model helps identify students' academic success or risk levels based on relevant academic factors.
Decision Tree Algorithm:
A decision tree is a supervised machine learning algorithm used for classification and regression tasks by splitting data into branches based on decision rules (Quinlan, 1986). It is widely used in predictive analytics due to its interpretability and effectiveness in handling complex datasets.
Educational Data Mining (EDM):
Educational Data Mining is the process of applying data analysis techniques to educational datasets to uncover useful insights about student learning patterns, behaviors, and academic performance (Romero & Ventura, 2010). It plays a crucial role in developing predictive models for student success.
Machine Learning:
Machine learning is a subset of artificial intelligence that enables computer systems to learn from data and make predictions without explicit programming (Mitchell, 1997). The decision tree model in this study uses machine learning techniques to analyze academic data and forecast student performance.
Student Risk Level:
Student risk level refers to the likelihood of a student facing academic challenges that may lead to poor performance or failure (Tinto, 1993). Identifying at-risk students early helps educators implement timely interventions to support their learning.
Educational Intervention:
Educational intervention involves strategic actions taken by teachers, administrators, or support systems to assist students in improving their academic outcomes (Slavin, 2002). The predictive model developed in this study aids in determining when interventions are necessary.
Feature selection:
Feature selection is the process of identifying the most relevant variables that contribute to the accuracy of a predictive model (Guyon & Elisseeff, 2003). In this study, factors like attendance, participation, and prior grades are selected as key features influencing academic performance.
…
