INTENSIF: Jurnal Ilmiah Penelitian dan Penerapan Teknologi Sistem Informasi
Not a member yet
174 research outputs found
Sort by
Enhancing Accessibility, Engagement, and Motivation in Counseling Services for Secondary Schools through Gamified Blended Mobile and Virtual Reality Therapy
Full text linkBackground: Secondary school counseling services often face challenges such as limited counselor availability and low student participation. Traditional counseling methods frequently fail to engage students, thus reducing both accessibility and impact. Integrating Virtual Reality (VR) and mobile-based interventions presents a promising solution to address these issues. Objective: This study aims to evaluate the effectiveness of a gamified blended mobile and VR therapy in enhancing accessibility, cognitive-emotional-behavioral engagement, and motivation within secondary school counseling services. Methods: A mixed-methods research design was employed, combining quantitative methods (pre- and post-intervention surveys, along with behavioral tracking) and qualitative methods (semi-structured interviews and thematic analysis of focus group discussions). These methods were chosen to capture both measurable impacts and participants’ perceptions of the intervention. A total of 384 students and 10 counselors participated in an 8-week intervention. Results: The intervention led to a significant improvement in the Accessibility Index (from 3.2 to 4.6). Additionally, engagement across cognitive, emotional, and behavioral dimensions showed marked improvements. Thematic analysis revealed that students appreciated the safety and realism provided by the digital counseling environment. Conclusion: The gamified blended therapy approach effectively enhanced counseling accessibility and multidimensional engagement, offering a scalable, student-centered solution for secondary school counseling services
Sentiment Analysis Accuracy for 2024 Indonesian Election Tweets Using CNN-LSTM With Genetic Algorithm Optimization
Full text linkBackground: The 2024 Indonesian Presidential Election is ideal for analyzing public sentiment on Twitter. Data collection began with crawling from the data source to create a dataset, which included 62,955 entries from Twitter, 126,673 entries from IndoNews, and a combined Tweet+IndoNews dataset totaling 189,628 entries. Objective: This study aims to explore sentiment using a hybrid model integrating Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) methods, with feature expansion via Word2Vec optimized by a Genetic Algorithm (GA). Methods: The research evaluates the effectiveness of the hybrid CNN-LSTM model in analyzing sentiment from 2024 Indonesian Presidential Election tweets, aiming for higher accuracy and deeper insights compared to traditional methods. Results: The hybrid CNN-LSTM model, optimized with a Genetic Algorithm, significantly enhances accuracy, achieving the highest accuracy of 84.78% for the news data, marking a 3.59% increase. Conclusion: This study illustrates the innovative application of a hybrid CNN-LSTM model with Word2Vec feature expansion and Genetic Algorithm optimization for sentiment analysis in a national election context, demonstrating how advanced techniques can improve accuracy and efficiency in sentiment analysis.
Enhancing Vision Transformer Performance with Rotation Based Augmentation for Classifying Images of Colon Cancer Pathology
Full text linkBackground: In medical imaging, classifying images of colon cancer pathology is still an essential challenge, especially for facilitating early diagnosis and successful intervention. Recently, Vision Transformer (ViT) models have demonstrated great promise for a variety of computer vision tasks, including the classification of medical images. However, the lack of annotated medical datasets and the intrinsic unpredictability of histopathology pictures sometimes restrict their performance. Objective: This study aims to enhance the performance of ViT models in colon cancer pathology classification by introducing a targeted data augmentation strategy, with a particular focus on rotation-based augmentation. Methods: We proposed a data augmentation pipeline that uses controlled changes to improve the number and diversity of training data. Like Rotation, Flip and Geometry are emphasized to replicate the real-world tissue orientation variations that are frequently seen in colon pathology slides. 10,000 JPEG pictures of colon cancer pathology, each with a resolution of 768 x 768 pixels, are used to train the models. We use models trained with and without the suggested augmentation pipeline to compare ViT performance across accuracy, sensitivity, and specificity in order to assess the impact of augmentation. Results: According to study results, rotation-based augmentation enhances ViT performance, achieving up to 99.30% accuracy and 99.50% sensitivity while preserving training times. In real-world pathology settings, where slide orientation varies greatly and can affect categorization consistency, these enhancements are especially pertinent. Conclusion: The proposed rotation-centric data augmentation technique enhances the performance of the ViT model in the classification of images showing colon cancer pathology
A Comparative Analysis of UTAUT and UTAUT 2 in M-Commerce and M-Banking
Full text linkBackground: The debate over the superiority between UTAUT and UTAUT 2 has driven the development of UTAUT 3. However, this latest model is still not the best solution, as several variables remain less influential. Objective: This study aims to determine which UTAUT model is superior in technology acceptance, particularly for m-commerce and m-banking. Methods: Multiple linear regression was used to determine the direction and magnitude of the influence of independent variables on the dependent variable, analyzing key factors affecting the adoption of mobile banking and m-commerce by comparing UTAUT and UTAUT 2. Results: In m-commerce, UTAUT highlights behavioral intention and facilitating conditions, while UTAUT 2 adds habit and price value as key influencing factors. Conclusion: For m-banking, both models are equally effective, but UTAUT 2 is superior due to the strong influence of habit. In m-commerce, UTAUT 2 is also preferable, as price value significantly affects behavioral intention
Understanding Student Acceptance of AI in Mojokerto Regency High Schools and a Framework for Effective Integration
Full text linkBackground: The use of AI in education is growing rapidly, especially in adaptive learning and automated feedback. Recent studies show widespread adoption of AI in higher education, but research at the secondary school level is limited. Factors such as ease of use, motivation, and institutional support play an important role accepting these technologies. Objective: The objective of this study is to investigate the acceptance and usage of the Question.AI application among high school students in Mojokerto Regency, to identify the factors that influence its adoption and effectiveness in enhancing learning outcomes. Methods: The methodology adopted for this research comprises a quantitative study design using a probability sampling method, specifically the Stratified Random Sampling technique. A total of 400 high school students from Mojokerto Regency participated. Data collection was conducted through structured questionnaires designed to evaluate factors influencing the adoption of the Question.AI application. Result: The result revealed that Facilitating Conditions (FC), Habit (H), and Hedonic Motivation (HM) significantly influence students\u27 behavioral intention to use the Question.AI application. Among these, Habit and Hedonic Motivation showed the strongest effect, indicating that students are more likely to adopt AI tools when their use becomes routine and satisfied. Conclusion: These results support the UTAUT2 framework and highlight the need for enjoyable user experiences and adequate support systems to drive sustained adoption. The findings contribute to understanding AI acceptance at the secondary education level and offer practical insights for integrating AI applications more effectively into school environments
Enhancing Multi-Class Classification of Non-Functional Requirements Using a BERT-DBN Hybrid Model
Full text linkBackground: Software requirements classification is essential to group Non-Functional Requirements (NFR) into several aspects, such as security, usability, performance, and operability. The main challenges in NFR classification are data limitations, text complexity, and high generalization needs. Objective: This research seeks to create a classification model using a hybrid of BERT and DBN, optimize hyperparameters, and improve data representation. Methods: A BERT and DBN-based approach is used, where DBN enhances BERT\u27s ability to extract hierarchical features. Bayesian Optimization determines the optimal hyperparameters and data augmentation is applied to enrich the dataset variation. The model is tested on the PROMISE dataset consisting of 625 data. Results: The BERT-DBN model achieves 95% accuracy on the baseline configuration and 94% on the extensive configuration, better than the previous model, BERT-CNN. The model shows stability without any indication of overfitting. Conclusion: The combination of data augmentation, hyperparameter optimization, and DBN\u27s ability to capture hierarchical patterns improves the accuracy of NFR classification, making it more effective than existing methods, and is expected to enhance text-based classification for software requirements
Enhancing the Decision Tree Algorithm to Improve Performance Across Various Datasets
Full text linkBackground: The Village Fund is an initiative by the central government to promote equitable regional development. However, it has also led to corruption. Many Indonesians share their opinions on the Village Fund on social media platforms like X, and news coverage is extensive on portals like detik.com. Objective: This study aims to classify data from social media and news coverage to enhance understanding. Methods: The research improves the decision tree algorithm by integrating other algorithms and techniques such as XGBoost and SMOTE. Ensuring high accuracy is vital for the credibility of machine learning classifications among the public. The study uses two different datasets, necessitating varied testing approaches. For the news portal dataset, a single test with seven labels is conducted, followed by enhancement with XGBoost. The X dataset undergoes two tests with datasets of 1200 and 3078 entries, using three labels. Conclusion: The evaluation results indicate that the highest accuracy achieved with the news portal data was 82%, thanks to a combination of decision tree algorithms with various parameters and the balancing effect of SMOTE. For the Twitter dataset with 3078 entries, the highest accuracy reached 95%, attributed to the application of ensemble techniques, particularly boosting
Performance of Deep Feed-Forward Neural Network Algorithm Based on Content-Based Filtering Approach
Full text linkBackground: Selecting a restaurant in a diverse city like Bandung can be challenging. This study leverages Twitter data and local restaurant information to develop an advanced recommendation system to improve decision-making. Objective: The system integrates content-based filtering (CBF) with deep feedforward neural network (DFF) classification to enhance the accuracy and relevance of restaurant recommendations. Methods: Data was sourced from Twitter and PergiKuliner, with restaurant-related tweets converted into rating values. The CBF combined Bag of Words (BoW) and cosine similarity, followed by DFF classification. SMOTE was applied during training to address data imbalance. Results: The initial evaluation of CBF showed a Mean Absolute Error (MAE) of 0.0614 and a Root Mean Square Error (RMSE) of 0.0934. The optimal DFF configuration in the first phase used two layers with 32/16 nodes, a dropout rate of 0.3, and a 20% test size. This setup achieved an accuracy of 81.08%, precision of 82.89%, recall of 76.93%, and f1-scores of 79.23%. In the second phase, the RMSprop optimizer improved accuracy to 81.30%, and tuning the learning rate to 0.0596 further increased accuracy to 89%, marking a 9.77% improvement. Conclusion: The research successfully developed a robust recommendation system, significantly improving restaurant recommendation accuracy in Bandung. The 9.77% accuracy increase highlights the importance of hyperparameter tuning. SMOTE also proved crucial in balancing the dataset, contributing to a well-rounded learning model. Future studies could explore additional contextual factors and experiment with recurrent or convolutional neural networks to enhance performance further
Comparative Analysis of Transformer-Based Method In A Question Answering System for Campus Orientation Guides
Full text linkThe campus introduction process is a stage where new students acquire information about the campus through a series of activities and interactions with existing students. However, the delivery of campus introduction information is still limited to conventional methods, such as using guidebooks. This limitation can result in students having a limited understanding of the information needed during their academic period. The one of solution for this case is to implement a deep learning system with knowledge-based foundations. This research aims to develop a Question Answering System (QAS) as a campus introduction guide by comparing two transformer methods, namely the RoBERTa and IndoBERT architectures. The dataset used is processed in the SQuAD format in the Indonesian language. The collected SQuAD dataset in the Indonesian language consists of 5046 annotated data. The result shows that IndoBERT outperforms RoBERTa with EM and F1-Score values of 81.17 and 91.32, respectively, surpassing RoBERTa with EM and F1-Score values of 79.53 and 90.18.The campus introduction process is a stage where new students acquire information about the campus through a series of activities and interactions with existing students. However, the delivery of campus introduction information is still limited to conventional methods, such as using guidebooks. This limitation can result in students having a limited understanding of the information needed during their academic period. The one of solution for this case is to implement a deep learning system with knowledge-based foundations. This research aims to develop a Question Answering System (QAS) as a campus introduction guide by comparing two transformer methods, namely the RoBERTa and IndoBERT architectures. The dataset used is processed in the SQuAD format in the Indonesian language. The collected SQuAD dataset in the Indonesian language consists of 5046 annotated data. The result shows that IndoBERT outperforms RoBERTa with EM and F1-Score values of 81.17 and 91.32, respectively, surpassing RoBERTa with EM and F1-Score values of 79.53 and 90.18
Technology Acceptance Analysis Using UTAUT: A Study of QRIS Acceptance during the Pandemic
Full text linkBackground: The COVID-19 pandemic situation has compelled society to practice physical distancing. One of the government\u27s efforts is to encourage the use of the QRIS payment method to minimize direct physical contact during transactions. Objective: The purpose of this research is to analyze the primary driving factors in the adoption of QRIS technology. The research urgency is to determine the most contributing predictor from the variables within the UTAUT model among the people of Jabodetabek. Methods: This research used the quantitative method by conducting an online survey among 384 respondents distributed across the Jabodetabek region. The sampling technique utilized was non-purposive sampling with criteria including domicile, age, reasons, frequency, and experience of QRIS usage. Conclusion: The results of the factor analysis test indicate that the performance expectancy and effort expectancy variables are combined into one variable, while the social influence variable is divided into two independent variables. The research findings reveal that the perceived risk variable is the predictor with the most significant contribution in the context of the pandemic situation. Future researchers are expected to be able to develop the research model in other contexts with different goals.Background: The COVID-19 pandemic situation has compelled society to practice physical distancing. One of the government\u27s efforts is to encourage the use of QRIS payment method to minimize direct physical contact during transactions. Objective: The purpose of this research is to analyze the primary driving factors in the adoption of QRIS technology. The research urgency is to determine the most contributing predictor from the variables within the UTAUT model among the people of Jabodetabek. Methods: This research used quantitative method by conducting an online survey among 384 respondents distributed across the Jabodetabek region. The sampling technique utilized was non-purposive sampling with criteria including domicile, age, reasons, frequency, and experience of QRIS usage. Conclusion: The results of the factor analysis test indicate that the performance expectancy and effort expectancy variables are combined into one variable, while the social influence variable is divided into two independent variables. The research finding reveal that the perceived risk variable is the predictor with the most significant contribution in the context of the pandemic situation. Future researches are expected to be able to develop the research model in the other context with different goals