cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
-
Contact Email
rji@mail.unnes.ac.id
Phone
-
Journal Mail Official
rji@mail.unnes.ac.id
Editorial Address
Sekaran, Kec. Gn. Pati, Kota Semarang, Jawa Tengah 50229
Location
Kota semarang,
Jawa tengah
INDONESIA
Recursive Journal of Informatics
Published by Universitas Negeri Semarang
ISSN : -     EISSN : 29866588     DOI : https://doi.org/10.15294/rji
Core Subject : Science,
Computer Science & IT
Recursive Journal of Informatics published by the Department of Computer Science, Universitas Negeri Semarang, a journal of Information Systems and Information Technology which includes scholarly writings on pure research and applied research in the field of information systems and information technology as well as a review-general review of the development of the theory, methods, and related applied sciences. We hereby invite friends to post articles and citation articles in our journals. We appreciate it if you would like to submit your paper for publication in RJI. The RJI publication period is carried out 2 periods in a year, namely in March and September.
Arjuna Subject : Ilmu Komputer (semua kategori) - Sistem Informasi
Articles 7 Documents
 1 
Search results for , issue "Vol. 3 No. 2 (2025): September 2025" : 7 Documents clear
Implementation of Bidirectional Long-Short Term Memory (Bi-LSTM) and Attention to Detect Political Fake News Using IndoBERT and GloVe Embedding Firmansyah, Adham Satria; Putra, Anggyi Trisnawan
Recursive Journal of Informatics Vol. 3 No. 2 (2025): September 2025
Publisher : Universitas Negeri Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15294/rji.v3i2.159

Abstract

Abstract. Indonesian political news is now increasingly spread through various media, especially social and online media. However, a lot of fake news are spread to bring down political opponents or attract public sympathy in order to find their own supporters. Of course, this news need to be watched out for and preventive measures must be taken so as not to cause misunderstanding in the wider community. Purpose: This study was conducted to detect the political news whether it’s classified as hoax or fact by its narration. Also, understanding how to build the news detector using corresponding architecture and word embeddings. Methods/Study design/approach: The model architecture of Bi-LSTM and attention mechanism is used to reach the goals from this study’s purposes. Many studies have been conducted to detect hoaxes but have not yet paid attention to the context of sentences and the contribution of words in a news text so that this architecture is made to overcome this problem. It uses IndoBERT to optimize the model for Indonesian language and also GloVe to obtain the word weights from pre-trained embedding. Then, the tokenization process is performed with IndoBERT and keras to generate token id and attention mask. After receiving the token id and attention mask as input, the data training process is performed for three architectural scenarios with each configuration of 20 epochs, batch size of 32, and the learning rate is 0.00001. Result/Findings: The results of this study are defined by a confusion matrix which contains accuracy, recall, precision, and F1-score as the evaluation. The combination of Bi-LSTM + Attention + IndoBERT + GloVe obtains the best result of 97,71% of accuracy, 96,33% of precision, 97,93% of recall, and 97,72% of F1-score. Novelty/Originality/Value: This study combines two word embeddings in order to make sure the weight of words is completely defined and optimized into the Bi-LSTM and attention mechanism architecture.
Brain Tumor Classification on Magnetic Resonance Imaging Images using Convolutional Neural Network with Cycle Generative Adversarial Network and Extreme Gradient Boosting Aji, Akbar Lintang; Sugiharti, Endang
Recursive Journal of Informatics Vol. 3 No. 2 (2025): September 2025
Publisher : Universitas Negeri Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15294/rji.v3i2.486

Abstract

Abstract. With the current advancement in technology, image classification process can be carried out through computer processing. This can also be applied to various fields, one of which is the health sector. The health sector is known for its high complexity in pattern recognizing of diseases. One of the diseases that is difficult to classify is brain tumors.             Purpose: This study aims to improve the accuracy of classification in brain MRI images, which are known to have a small and unbalanced sample. This limitation poses challenges in developing an effective classification model. The classification model is highly dependent on the quantity of data used for training. Therefore, data augmentation techniques play a crucial role in influencing the model's performance. Methods/Study design/approach: In this study, CNN model using VGG-19 architecture was used to learn feature of brain tumor in brain MRI images. Additionally, CycleGAN is used to augment and balance the data, addressing issues related to data scarcity and imbalance, thus improving diversity of the dataset. And then, XGBoost is applied to classify the feature learned by the CNN model. Result/Findings: CycleGAN has the ability to generate new image by transferring characteristics between images with different classes, making it a suitable to replace traditional data augmentation techniques in CNN. Additionally, XGBoost can be used to improve the classification results by classifying the features learned by CNN model during the training process. The proposed combination method achieves a highest accuracy of 97.37%. Novelty/Originality/Value: CNN combined with CycleGAN and XGBoost successfully improved the accuracy of the model and addresed data scarcity and imbalances in the dataset used. This combined method can improve the accuracy of the classification models. This is proven by an accuracy increase of 0.36% when compared to previous research.
Selection of Trading Indicators Using Machine Learning and Stock Close Price Prediction with the Long Short Term Memory Method Rafli, Alfandy Himawan Bagus; Purwinarko, Aji
Recursive Journal of Informatics Vol. 3 No. 2 (2025): September 2025
Publisher : Universitas Negeri Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15294/rji.v3i2.945

Abstract

Abstract. Humans have a limit to their physical ability to work, so investment is needed to meet their needs and other goals according to their wants and needs. Investment has many types and risks according to the portion of the return value, such as mutual funds, bonds and stocks. Stocks are a form of investment that has a high risk because of the rapid fluctuations in stock values. Prediction of stock movements is usually assisted by indicators, but predictions using indicators require complex analysis because of the diverse periods and different movements in each stock data case.  Purpose: To predict the closing price of BBCA and BBRI shares in the next 10 days by considering the count of technical indicators in the form of Moving average (MA), Exponential moving average (EMA), Rate Of Change (ROC), Price Momentum, Relative Strength Index (RSI), Stochastic Oscillator in periods 21, 63 and 252. Methods/Study design/approach: This research was conducted by comparing the accuracy of Random Forest, Decision Tree, KNN, SVM using K-fold Cross Validation then the method with the best accuracy was used to find out how much velue from the trading indicators used and predict the closing price of shares per day at BBRI and BBCA companies for the next 10 day period using the LSTM algorithm. Result/Findings: The best accuracy in the k-fold cross validation process is random forest. random forest is used to train indicator data in determining 5 indicators along with the period that has the highest value, in this test it produces values on BBCA data in order, namely ROC63, RSI63, MOM63, MA252, EMA21 while on BBRI data in order, namely ROC63, MOM63, RSI63, MA252, MA21. This indicator is used in the price forecasting process with the LSTM method to determine the closing price in the next 10 days. The LSTM method in this study resulted in 96.8% accuracy for BBCA and 96.4% accuracy for BBRI. Novelty/Originality/Value: The forecasting accuracy on BBCA is 96.8% and the forecasting accuracy on BBRI is 96.4%. This shows that the accuracy results are classified as good because the prediction results are close to the actual results. The data training process is expected to help traders in making stock buying and selling decisions that are adjusted to the fundamental aspects of the company.
Implementation of the Term Frequency-Inverse Document Frequency Method for Mental Health Classification Using Algorithm Support Vector Machine Minatika, Ilfa; Arifudin, Riza
Recursive Journal of Informatics Vol. 3 No. 2 (2025): September 2025
Publisher : Universitas Negeri Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15294/rji.v3i2.1921

Abstract

Abstract. Mental health is a person's emotional, social and psychological condition. A person's mental health level can be influenced by emotional experiences, behavior, environment and family educational background. A person's psychological well-being can be influenced by a person's behavior, where they live, the education they receive, and their emotional experiences. It is important not to underestimate the existence of mental health disorders because the number of cases is currently increasing. Purpose: Using SVM algorithm and TF-IDF method can produce good accuracy for classification text. Therefore this research aims to determine the implementation of the use of the TF-IDF method and the SVM algorithm in mental health classification and to determine the accuracy results of using these methods. Study Method/Design/Approach: The methods used in the research this for the mental health classification is Term Frequency-Inverse Document Frequency used in the vectorization process to convert text into a numerical representation, as well as using the Support Vector Machine algorithm in modeling. The dataset used is the Mental Health Corpus dataset obtained from the Kaggle website. This dataset consists of two classes containing text and labels totaling 27,977 data. Before applying the model, preprocessing is carried out first, namely cleaning the text using stopword removal and stemming. After cleaning the text, the next process is vectorization using CountVectorizer and TF-IDF. Results/Findings: In this study the SVM algorithm was used four kernels, namely the linear kernel, the RBF kernel, the polynomial kernel, and the later sigmoid kernel get the best accuracy results on the RBF kernel if compared to with other kernels. Accuracy results obtained _ of 92.62%, value precision of 92.64%, value recall 92.62%, and value f1-score 92.62%. Novelty/Originality/Value: So, it can be concluded that the application of the SVM algorithm and the TF-IDF method is possible used for classification mental health results mark high accuracy.
Implementation of the K-Nearest Neighbor Algorithm (KNN) with Principal Component Analysis to Diagnose Tuberculosis Putri, Yuliana; Alamsyah, Alamsyah
Recursive Journal of Informatics Vol. 3 No. 2 (2025): September 2025
Publisher : Universitas Negeri Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15294/rji.v3i2.5235

Abstract

Purpose: Tuberculosis (TB) is an infectious disease that attacks the respiratory organs, the lungs, and some can attack organs outside the lungs. Indonesia is one of the largest contributors to TB cases with around 320,000 new cases every year. Delays in diagnosing TB disease can cause a higher number of deaths due to errors in the treatment of sufferers. This makes the early diagnosis of TB disease important as early as possible. The research carried out aims to implement machine learning techniques to help diagnose TB disease. Methods: The research was carried out using the K-Nearest Neighbor (KNN) classification algorithm which was optimized with the Principal Component Analysis (PCA) feature selection technique. The dataset used consists of 577 data with 12 attributes labeled patients with tuberculosis and patients who do not have tuberculosis. Result: From the research that has been conducted, models that implement the KNN algorithm with PCA produce models with better performance than models that only implement KNN. The model that only uses KNN gets an accuracy of 92.528%, while the model that uses KNN and PCA gets an accuracy of 98.85%. This shows that the implementation of KNN and PCA is able to produce a good tuberculosis diagnosis model and can be used to assist in the early diagnosis of tuberculosis. Novelty: Using PCA in the feature selection process can reduce unnecessary attributes. It is a PCA that helps reduce the dimensionality, simplifies the visualization and interpretation of complex data sets. The use of PCA has been proven to be able to optimize the performance of the KNN algorithm for the detection of tuberculosis.  
Sentiment Analysis of Jobstreet Application Reviews on Google Play Store Using Support Vector Machine Algorithm with Adaptive Synthetic Shantika, Febryan Surya; Abidin, Zaenal
Recursive Journal of Informatics Vol. 3 No. 2 (2025): September 2025
Publisher : Universitas Negeri Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15294/rji.v3i2.11891

Abstract

Abstract. Purpose: This research aims to test the performance result of the Support Vector Machine (SVM) classification algorithm using the help of Adaptive Synthetic (ADASYN) oversampling to analyze sentiment in Jobstreet application reviews on the Google Play Store. Sentiment analysis is a significant method to understand the market needs and application improvement. Methods/Study design/approach: The dataset originates from Google Play reviews gained using the scrapping method, comprising 5,174 reviews with 11 attributes. The process begins with data scrapping, data labeling, and data preprocessing, including casefolding, tokenizing, filtering, and stemming using Python programs. The data is then weighted and split using an 80:20 ratio. Then applying oversampling ADASYN on a clean dataset before using SVM classification to produce the performance result. Result/Findings: Both scenarios are conducted on SVM classification to classify the dataset. The evaluation results indicate that using SVM classification without ADASYN produces an accuracy result of 89.08%. Other scenarios by using SVM classification with the ADASYN sampling approach produce an accuracy result of 89.95%. The performance in accuracy result by using the ADASYN sampling approach on SVM classification shows an increasing result of 0.87%. Novelty/Originality/Value: This study employs two result scenarios of SVM classification by using the ADASYN sampling approach. It contributes to the literature by demonstrating the usability of the ADASYN oversampling approach to optimalize the SVM classification result used for sentiment analysis in Jobstreet application reviews on the Google Play Store.
Improving Pantun Generator Performance with Fine Tuning Generative Pre-Trained Transformers Sodikkun, Achmat; Budiman, Kholiq
Recursive Journal of Informatics Vol. 3 No. 2 (2025): September 2025
Publisher : Universitas Negeri Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.15294/ge6xey51

Abstract

Purpose: The study aims to address the challenges in generating high-quality pantun, an important element of Indonesian cultural heritage. Traditional methods struggle with limited vocabulary, variation, and consistency in rhyme patterns. This research seeks to enhance the performance of a pantun generator by applying fine-tuning techniques to the Generative Pre-trained Transformers (GPT) model, coupled with post-processing, and validated by linguistic experts. Methods/Study design/approach: The research involves fine-tuning the GPT model using a dataset of Indonesian pantun. The methodology includes dataset collection, data pre-processing for cleaning and adjustment, and hyperparameter optimization. The effectiveness of the model is evaluated using perplexity and rhyme accuracy metrics. The study also incorporates post-processing to refine the generated pantun further. Result/Findings: The study achieved a best perplexity value of 14.64, indicating a strong predictive performance by the model. Post-processing significantly improved the rhyme accuracy of the generated pantun to 89%, a substantial improvement over previous studies by Siallagan and Alfina, which only achieved 50%. These results demonstrate that fine-tuning the GPT model, supported by appropriate hyperparameter settings and post-processing techniques, effectively enhances the quality of generated pantun. Novelty/Originality/Value: This research contributes to the development of generative applications in Indonesian, particularly in the context of cultural preservation. The findings highlight the potential of fine-tuning GPT models to improve language generation tasks and provide valuable insights for creative and educational applications. The validation by experts ensures that the generated pantun adheres to established writing standards

Page 1 of 1 | Total Record : 7

 1 

Filter by Year

2025 2025


Reset
Filter By Issues
All Issue Vol. 3 No. 2 (2025): September 2025 Vol. 3 No. 1 (2025): March 2025 Vol. 2 No. 2 (2024): September 2024 Vol. 2 No. 1 (2024): March 2024 Vol. 1 No. 2 (2023): September 2023 Vol. 1 No. 1 (2023): March 2023