Chapter II

REVIEW OF RELATED SYSTEMS

The advancement of technology has significantly influenced educational

institutions by integrating machine learning, data visualization, and web-based platforms

to enhance student guidance and information dissemination. This chapter reviews

related systems and research studies that contribute to the development of the Arteche

National High School Website with Machine Learning and Data Visualization. The

studies presented focus on machine learning applications in strand recommendation,

school information systems, and automated exam analysis. These references serve as a

foundation for understanding the potential impact of technology-driven decision-making

tools in educational settings.

Atienza et al. (2022) developed a web-based career track recommender system

aimed at assisting guidance counselors in identifying the most suitable senior high

school track for students. Utilizing Deep Neural Networks (DNN), the system analyzed

academic grades and socio-demographic factors, achieving 83.11% accuracy in strand

prediction. Their findings emphasize how AI-powered counseling enhances decision-

making in education.

To improve career recommendations, Smith and Rodriguez (2023) introduced a

Hybrid Recommender System that combined Collaborative Filtering and Content-Based

Filtering. Their model resulted in a 12% increase in accuracy compared to traditional

track selection methods, proving the effectiveness of hybrid recommendation

approaches.
 Brown et al. (2024) explored machine learning techniques such as Support

Vector Machines (SVM) and Decision Trees for predicting suitable academic strands.

Their findings revealed that Decision Trees outperformed SVM when handling

categorical student data, achieving 85% classification accuracy.

Advancing predictive models, Bhushan et al. (2024) implemented Artificial Neural

Networks (ANN) with Principal Component Analysis (PCA) to forecast Semester Grade

Point Average (SGPA). Their research demonstrated that ANN with PCA surpassed

traditional models like Random Forest and Logistic Regression, achieving 86.3%

accuracy.

A similar focus on student outcomes is evident in Lopez and Wang (2023), who

designed a dropout prediction system using Random Forest and XGBoost. Their results

highlighted that ensemble models provided 89% accuracy in identifying students at risk

of dropping out, enabling early intervention programs.

By examining academic performance trends, Garcia et al. (2022) developed a

predictive grading model utilizing Naïve Bayes and K-Nearest Neighbors (KNN). Their

study demonstrated that KNN performed exceptionally well in handling complex grading

patterns, achieving 84% prediction accuracy.

The significance of data visualization (DV) in education is emphasized by

Suganthalakshmi and Saravanakumar (2024), who demonstrated how interactive

dashboards enhance data interpretation and academic insights for students and

educators.

Focusing on academic monitoring, Wilson (2023) examined the impact of

interactive dashboards and discovered that schools implementing real-time analytics

observed a 20% improvement in student participation and overall monitoring efficiency.

School Information and Publication System designed to disseminate announcements,

updates, and event schedules. The system received an "Excellent" usability rating,

proving its effectiveness in improving school information accessibility.

Liu et al. (2024) investigated the efficiency of automated school management

platforms, reporting that AI-driven administrative systems reduced manual workload by

60%, leading to more efficient school operations.

Luo (2024) explored machine learning in student decision-making, comparing

models like Decision Trees, SVM, ANN, and Random Forests. Their analysis showed

that Random Forest was the most effective, excelling in handling multi-dimensional

educational data.

Further improving student performance forecasting, Chowdhury and Singh

(2023) developed a Long Short-Term Memory (LSTM) network model, which

outperformed traditional forecasting methods by 18%, proving the value of deep learning

in trend prediction.

Neural networks have also been applied in risk detection. Anderson et al. (2022)

examined predictive analytics for identifying at-risk students, concluding that neural

networks achieved 89% accuracy in detecting students likely to fail, reinforcing the

importance of early intervention programs.

A different approach to educational analysis was explored by Williams and Carter

(2023), who utilized Natural Language Processing (NLP) to analyze student feedback.

Their system successfully automated sentiment analysis, improving school feedback

evaluation and increasing student satisfaction by 22%.

learning-based recommendation system. Their study revealed that AI-driven

recommendations led to improved course selection outcomes and higher student

satisfaction.

Personalized tutoring systems have also been explored. Gonzalez and Lee

(2022) developed an AI-based tutoring system using deep reinforcement learning,

resulting in an average test score improvement of 15% for students who used the

system regularly.

Career prediction remains a relevant topic in education research. Fernandez et

al. (2024) compared Bayesian Networks and Decision Trees in career track

recommendations, concluding that Bayesian Networks provided higher interpretability

and accuracy.

Expanding on early academic risk detection, Santos and Kim (2023) developed a

neural network-based early warning system, which significantly improved the

identification of students at risk of failing their courses.

A different aspect of adaptive learning was examined by Johnson et al. (2022),

who implemented Recurrent Neural Networks (RNNs) to personalize learning

experiences. Their results showed that adaptive learning systems increased student

engagement and retention rates.

Conversational AI has also played a role in education. Harris and Nguyen (2023)

developed a smart academic advisor chatbot that utilized NLP and deep learning, which

boosted student engagement in academic advising by 30%.

model and found that Random Forest achieved 91% accuracy in identifying students

inclined toward STEM fields, assisting schools in targeted career guidance.

Mendoza et al. (2023) introduced a student profiling system using k-Means

clustering, which effectively categorized students based on academic performance,

learning styles, and interests.

By analyzing study habits, Park and Chen (2022) applied association rule mining,

uncovering strong correlations between consistent study routines and long-term

academic success.

Ramirez and Foster (2023) explored key predictors of student success, using

factor analysis to determine that extracurricular involvement and study habits were

significant indicators of academic achievement.

To support career decision-making, Nelson and White (2024) created an AI-

based career assessment tool, which matched students with careers based on their

skills and preferences, achieving an 87% accuracy rate.

Bennett and Yoon (2023) focused on learning analytics for personalized

education, demonstrating that data-driven customization improved learning outcomes by

20%.

Kumar and Zhang (2024) examined machine learning in scholarship eligibility

prediction, showing that XGBoost models provided better fairness and accuracy

compared to traditional statistical methods.

A unique perspective on student engagement was provided by Singh et al.

(2023), who used social media data mining to analyze student behavior. Their study
found that social media activity correlated with academic motivation and engagement

levels.

The reviewed literature collectively underscores the impact of AI, machine

learning, and data visualization in enhancing education systems. Several studies, such

as those by Atienza et al. (2022), Smith and Rodriguez (2023), and Fernandez et al.

(2024), focused on career track recommendations, while others, including Bhushan et al.

(2024), Lopez & Wang (2023), and Garcia et al. (2022), highlighted predictive analytics

for academic performance and dropout prevention.

Additionally, Wilson (2023) and Suganthalakshmi & Saravanakumar (2024)

demonstrated how data visualization enhances decision-making, whereas Caratiquit

(2021) and Liu et al. (2024) provided insights into school information systems improving

administrative processes.

Despite these advancements, none of the existing research fully integrates AI-

driven strand recommendations, real-time data visualization, and school information

management in a single platform. The Arteche National High School Website seeks to

bridge this gap by developing an all-in-one solution that leverages machine learning for

academic guidance while incorporating interactive data visualization to enhance

decision-making and school communication.