Claim Missing Document
Check
Articles

Found 4 Documents
Search
Journal : Building of Informatics, Technology and Science

Comparative Assessment of Low Job Competitiveness Among University Graduates Using Naïve Bayes and KNN Algorithms Hamonangan, Ricardo; Palupi, Irma; Gunawan, Putu Harry
Building of Informatics, Technology and Science (BITS) Vol 6 No 1 (2024): June 2024
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i1.5406

Abstract

Tracer studies investigate the career outcomes of graduates, encompassing job search experiences, employment conditions, and the application of acquired skills post-graduation. These studies are pivotal for universities and colleges to assess graduate success and shape educational policies. This study aims to elucidate the factors contributing to low job competitiveness through the application of classification models like KNN and Naïve Bayes. It also evaluates how competencies developed during university studies impact this scenario. Key issues addressed include the identification of factors causing low job competitiveness and the assessment of competencies trained during university education. Utilizing a dataset comprising two classes and seven features, the KNN method achieved an accuracy of 71.00%, while Naïve Bayes achieved 70.00%. The data set size is 1853 (around 20% of the survey sample) of unemployed alumni. The results indicate that the lack of specific competencies, particularly those related to practical skills and real-world application, is a major factor contributing to low job competitiveness. The results highlight a specific competency as most crucial in the KNN model, whereas different competencies play significant roles in the Naïve Bayes model. Despite variations in competency importance across models, all features significantly contribute to predictions. This research enhances the classification of workforce competitiveness levels within tracer studies and underscores the potential of KNN and Naïve Bayes algorithms to identify factors influencing low job competitiveness. These findings support informed decision-making in academic and career development initiatives, emphasizing the critical influence of university-trained competencies on job market readiness.
Decision Tree Algorithm for Predicting Alumni Job Competitiveness Through Waiting Time Working Panuluh, Bagus; Palupi, Irma; Gunawan, Putu Harry
Building of Informatics, Technology and Science (BITS) Vol 6 No 2 (2024): September 2024
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i2.5647

Abstract

The absorption of alumni from universities into the world of work is an essential indicator that universities must pay attention to. One-way universities can pay attention to their alums is through tracer studies, where they can evaluate their curriculum's relevance to what is needed in today's world of work. One aspect that can be seen from the tracer study to assess the competitiveness of alums is the waiting time for alums to get their first job. This is because the sooner alums get jobs, the better the curriculum the university provides to students. This research aims to apply machine learning to predict the waiting time for alums from Telkom University to get their first job and find out what factors influence the waiting time for work. The algorithm used in the research is the Decision Tree with hyperparameter tuning using Grid Search and feature selection application. There are 3 methods of feature selection used for comparison: Spearman's Rank Correlation, Chi-square, and Principal Component Analysis. This research produces the best prediction model in applying Chi-square and hyperparameter tuning with an accuracy of 0.79, recall of 0.79, precision of 0.80, and F1-Score 0.75. Several features, such as the number of companies registered, how to find and get work, internship and practicum experience, ethical competency, discussion, and IT skills, have the biggest effects on the model.
Predicting University Graduates Employability Using Support Vector Machine Classification Haikal, Muhamad Fachri; Palupi, Irma
Building of Informatics, Technology and Science (BITS) Vol 6 No 2 (2024): September 2024
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v6i2.5655

Abstract

The absorption of graduates into the world of work is a key indicator of higher education institution success, especially amid the tight job market competition due to increasing graduate numbers. Understanding employability and the factors that influence it is crucial for higher education institution to enhance education quality and facilitate graduates' transitions to employment. This research aimed to predict the employability of Telkom University students through their initial job income. Methods involved feature manipulation techniques like Principal Component Analysis, Spearman's rank correlation, and the Chi-square test of independence, followed by SMOTE-ENN to address data imbalance. Modeling was conducted using a Support Vector Machine with Randomized Search hyperparameter tuning, analyzed through Permutation Feature Importance to identify factors affecting employability. The result showed the enhanced SVM model with SMOTE-ENN, Spearman’s rank correlation coefficient as feature selection and randomized search hyperparameter tuning achieved the highest precision, recall, f-score, and accuracy of approximately 0.70, 0.73, 0.71, and 0.73, respectively. Competency features such as ethics, english skills, IT skills, and knowledge were identified as the most influential factors.
Mental Health Sentiment Analysis on Twitter using Ensemble Learning Algorithm Aziz, Kemal; Wahyudi, Bambang Ari; Palupi, Irma
Building of Informatics, Technology and Science (BITS) Vol 7 No 2 (2025): September 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i2.7763

Abstract

Mental health problems have become an important health issue around the world. Poor understanding as well as low mental health awareness contribute to mental health healing efforts. In particular, Social media is becoming a platform for people to convey feelings and emotions. A dataset of 20,000 English tweets, equally divided into 10,000 depressed and 10,000 non-depressed tweets, which were cleaned and processed using Term Frequency-Inverse Document Frequency (TF-IDF) for feature extraction. The method used in this sentiment analysis introduces an ensemble learning framework that combines Naïve Bayes, Support Vector Machine, and Random Forest classifiers, using majority voting for prediction. Each classifier was optimized using the best parameters, and the models were validated through 5-fold cross-validation. The experimental results show that Naïve Bayes with α = 1 achieved an accuracy of 76.23% while Random Forest with 5000 trees at 76.77%, and Support Vector Machine with a linear kernel at 75.32%. By combining these classifiers, the ensemble model reached the highest accuracy of 77.88%, demonstrating the effectiveness of combining multiple models to improve performance.