Claim Missing Document
Check
Articles

Found 4 Documents
Search

Determining community happiness index with transformers and attention-based deep learning Wicaksana, Hilman Singgih; Kusumaningrum, Retno; Gernowo, Rahmat
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 13, No 2: June 2024
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v13.i2.pp1753-1761

Abstract

In the current digital era, evaluating the quality of people's lives and their happiness index is closely related to their expressions and opinions on Twitter social media. Measuring population welfare goes beyond monetary aspects, focusing more on subjective well-being, and sentiment analysis helps evaluate people's perceptions of happiness aspects. Aspect-based sentiment analysis (ABSA) effectively identifies sentiments on predetermined aspects. The previous study has used Word-to-Vector (Word2Vec) and long short-term memory (LSTM) methods with or without attention mechanism (AM) to solve ABSA cases. However, the problem with the previous study is that Word2Vec has the disadvantage of being unable to handle the context of words in a sentence. Therefore, this study will address the problem with bidirectional encoder representations from transformers (BERT), which has the advantage of performing bidirectional training. Bayesian optimization as a hyperparameter tuning technique is used to find the best combination of parameters during the training process. Here we show that BERT-LSTM-AM outperforms the Word2Vec-LSTM-AM model in predicting aspect and sentiment. Furthermore, we found that BERT is the best state-of-the-art embedding technique for representing words in a sentence. Our results demonstrate how BERT as an embedding technique can significantly improve the model performance over Word2Vec.
Penerapan Word2Vec dan SVM dengan Hyperparameter Tuning untuk Deteksi Phishing Wicaksana, Hilman Singgih; Huda, Khairul
JURIKOM (Jurnal Riset Komputer) Vol 12, No 3 (2025): Juni 2025
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/jurikom.v12i3.8729

Abstract

The advancement of information technology in today’s digital age takes place very rapidly from one time to another. This phenomenon is accompanied by increasing cybersecurity threats like phishing. Phishing links are often designed with uniform resource locator (URL) structures that appear convincing and are difficult to distinguish from genuine links. This research proposes a word-to-vector (Word2Vec) and Support Vector Machine (SVM) approach with hyperparameter tuning where Word2Vec is a word embedding technique used to create a word vector representation of a particular URL, SVM is used as a machine learning (ML) approach used in this research, and hyperparameter tuning is used as a technique to find the best combination of parameters to produce an optimal SVM model in detecting phishing. The purpose of this research is to compare the performance between SVM and XGBoost models that have been optimized and deploy ML models into a prediction system using the Streamlit framework to detect phishing based on input made by users in the form of certain URLs. The findings of this study indicated that the SVM model performed very well compared to the XGBoost model, with precision, recall, f1-score, and accuracy values of about 99.84% for SVM. On the other hand, the XGBoost model recorded precision, recall, f1-score, and accuracy values of about 99.70% each. Thus, the SVM model is the optimal model to detect phishing precisely and accurately.
Penerapan Word2Vec dan SVM dengan Hyperparameter Tuning untuk Deteksi Phishing Wicaksana, Hilman Singgih; Huda, Khairul
JURNAL RISET KOMPUTER (JURIKOM) Vol. 12 No. 3 (2025): Juni 2025
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/jurikom.v12i3.8729

Abstract

The advancement of information technology in today’s digital age takes place very rapidly from one time to another. This phenomenon is accompanied by increasing cybersecurity threats like phishing. Phishing links are often designed with uniform resource locator (URL) structures that appear convincing and are difficult to distinguish from genuine links. This research proposes a word-to-vector (Word2Vec) and Support Vector Machine (SVM) approach with hyperparameter tuning where Word2Vec is a word embedding technique used to create a word vector representation of a particular URL, SVM is used as a machine learning (ML) approach used in this research, and hyperparameter tuning is used as a technique to find the best combination of parameters to produce an optimal SVM model in detecting phishing. The purpose of this research is to compare the performance between SVM and XGBoost models that have been optimized and deploy ML models into a prediction system using the Streamlit framework to detect phishing based on input made by users in the form of certain URLs. The findings of this study indicated that the SVM model performed very well compared to the XGBoost model, with precision, recall, f1-score, and accuracy values of about 99.84% for SVM. On the other hand, the XGBoost model recorded precision, recall, f1-score, and accuracy values of about 99.70% each. Thus, the SVM model is the optimal model to detect phishing precisely and accurately.
Optimized Machine Learning Approach for Malware Detection Wicaksana, Hilman Singgih; Huda, Khairul
Jurnal Sisfokom (Sistem Informasi dan Komputer) Vol. 15 No. 01 (2026): JANUARY
Publisher : ISB Atma Luhur

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32736/sisfokom.v15i01.2547

Abstract

The rapid evolution of information technology has created vast opportunities in multiple domains, yet it also brings critical challenges in the realm of cybersecurity, particularly with the growing frequency of malware attacks. Modern malware utilizes advanced evasion and spreading techniques, such as polymorphic and metamorphic transformations, which undermine the performance of conventional detection systems. This research aims to evaluate and compare the effectiveness of several machine learning algorithms optimized through hyperparameter tuning to determine the most accurate and reliable model for malware detection. The study applies a supervised learning approach using labeled data and examines five algorithms: Multilayer Perceptron, Random Forest, Support Vector Machine, Extreme Gradient Boosting, and Hist Gradient Boosting. Each model was fine-tuned to identify its optimal configuration, and performance was measured using accuracy, precision, recall, and F1-score. The experiments were conducted on a dataset comprising 58,596 records that had been thoroughly cleaned and preprocessed. The findings indicate that the Multilayer Perceptron achieved superior results, obtaining 99.97% across all evaluation metrics. These outcomes demonstrate the model’s strong potential for reliable malware detection and its suitability for integration into cybersecurity frameworks that demand fast response, high precision, and adaptability to evolving attack patterns.