Journal of Dinda : Data Science, Information Technology, and Data Analytics

Design Of A Decision Support System For The Graduation Of New Student Candidates Based On MVC

Fivy Nur Safitri, Daniel Yeri Kristiyanto — Tue, 13 Feb 2024 00:00:00 +0800

The selection process for new students in the field of education is crucial and warrants careful consideration. IT Telkom Purwokerto has a dedicated division, the Admission Unit, responsible for the selection of new students. However, this process often encounters errors, such as miscalculations in the average scores of three subjects, discrepancies between new student data and graduation guideline data, and prolonged simulation processes for graduation. This study proposes a solution to these issues through the implementation of an MVC-based Decision Support System (DSS) for determining the eligibility of new student admissions. The Prototype methodology was chosen to develop an MVC-based system as a resolution to these issues. The criteria used in this research to determine new student admissions involve various factors, including the chosen high school major, interest in the offered majors, average mathematics scores, and the average scores of three main subjects: mathematics, Bahasa Indonesia, and English. The outcomes of this research include the development of an MVC-based decision support system that aims to determine the admission status of new students. It is anticipated that the implementation of this decision support system based MVC will not only aid relevant personnel in the admission decision process but also mitigate potential issues that may arise. The research contributes to the enhancement of the efficiency and accuracy of the new student selection process at IT Telkom Purwokerto.

Random Forest Machine Learning for Spam Email Classification

Rizky Ageng, Rafdhani Faisal, Solahuddin Ihsan — Thu, 15 Feb 2024 00:00:00 +0800

This research discusses the crucial role of email as a main element in digital communication, facilitating information transfer and serving as an advertising platform. However, the problem of email spam, which involves sending unsolicited commercial messages, has had negative impacts such as consuming large amounts of resources and disrupting user experience. With its affordable cost and ease of sending messages to thousands of recipients, email spam includes product promotions, pornographic material, viruses and irrelevant content. The impact includes loss of time and damage to the user's computer resources. To address this problem, email services provide advanced spam filters that use email content analysis and machine learning techniques. This research focuses on the use of the Random Forest Classification algorithm as a basis for filtering spam emails. Although Random Forest is known to have strong classification capabilities, the risk of overfitting is a challenge. Therefore, this study adopts the Randomized Search CV method to identify the best parameter combination, ensuring the reliability of the model in dealing with the complexity of diverse email datasets. With this approach, this research contributes to the development of effective solutions to reduce the impact of email spam in digital communications.

Prediction of Obesity Classification Using K-Means Clustering

Aditya Wildan, Helmy Akmal Burhansyah, Choki Ferdiansyah — Thu, 15 Feb 2024 00:00:00 +0800

This paper aims to determine the difference between someone who is obese and who is not and classify the level of obesity by utilizing the K-Means clustering algorithm to group them. The move was taken as part of obesity prevention efforts, with the hope that a deeper understanding of the distribution of obesity within specific categories could help design more specific and effective interventions. Using this approach, it is hoped that this study can contribute to our understanding of the complexities of obesity and encourage more precise and targeted preventive measures. In this study we used datasets from Kaggle. It is used to classify the difference between underweight and overweight people. In this study, data was processed using Data Mining techniques with the K-Means method. Based on the classification, four clusters were categorized. Cluster 0 in this cluster only has women, with an age range ranging from 45 to 60 years. Relatively thin to normal weight. Cluster 1 only has men, with an age range of more than 40 years and 55 to 60 years. People in this cluster are overweight or obese. Cluster 2 women aged 15-70 years make up the majority in this group, with women aged 55-60 years as the highest proportion. In general, they have a normal weight. Many underweight individuals aged 10-45 years, with the highest proportion at the age of 20-25 years. The classification results show that men have a higher likelihood of suffering from obesity than women. Therefore, obesity prevention needs to be done, one of which is by applying a healthy lifestyle.

K-Means Clustering Algorithm: A Study on Unemployment Rates in Districts/Cities in Three Highest Provinces

Mohammad Dian Purnama, Mutia Eva Mustafidah — Thu, 22 Feb 2024 00:00:00 +0800

Unemployment is a recurring issue every year, particularly in provinces with high unemployment rates, posing economic and social challenges. West Java, Riau Islands, and Banten are identified as the three provinces with the highest unemployment rates, exceeding 8% in the year 2022. Hence, this study aims to delve into the unemployment scenario in these provinces, considering various influencing factors drawn from relevant previous research. The primary objective of this research is to obtain the classification results of regencies/cities in West Java, Riau Islands, and Banten based on unemployment indicators. The findings reveal four clusters: Cluster 1 comprises 13 regencies/cities with the lowest unemployment rates, Cluster 2 includes 4 regencies/cities with low unemployment rates, Cluster 3 consists of 13 regencies/cities with moderate unemployment rates, and Cluster 4 encompasses 12 regencies/cities with high unemployment rates.

Identifying Fake News Using Long-Short Term Memory Model

Farhan Wundari, Muhammad Nathan Asy Syaiba Amien, Dida Haiman Irtsa — Mon, 04 Mar 2024 11:03:56 +0800

Designed to deceive readers and manipulate public opinion, fake news can be created for a variety of reasons ranging from political propaganda to generating revenue through clickbait. Another significant challenge in combating fake news is the difficult balance between curbing misinformation and preserving free speech, though some argue for stricter regulations to control the spread of fake news. Thus, the purpose of this study is to identify fake news using Long-Short Term Memory (LSTM). LSTM models are often used to analyze the linguistic features of news articles or social media posts. The dataset we used comes from a dataset of fake news on Kaggle's website. The proposed method can identify fake news with average precision, recall, accuracy, and f-measure values of 0.94, 0.96, 0.94, and 0.95. The results showed that LSTM provides superior performance compared to the Support Vector Classifier, Logistic Regression, and Multinomial Naive Bayes methods.