CommIT (Communication and Information Technology) Journal

The Comparison of Deep Learning Models for Indonesian Political Hoax News Detection

Oktavia Citra Resmi Rachmawati; Zakha Maisat Eka Darmawan — Wed, 21 Aug 2024 00:00:00 +0000

Indonesia is the world’s fourth most populous country and has a diverse sociopolitical landscape. Political fake news exacerbates existing social divisions and causes political polarization in Indonesian society. Hence, studying it as a specific challenge can contribute to broader discussions on the impact of fake news in different contexts. The researchers propose a hoax news detection system by developing a deep learning model with various lapses against a data set preprocessed using term-frequency and token filtering to represent the most prominent words in each class. The researchers compare the layers with the potential to have high performance in predicting the falsity of Indonesian political news data by observing the models based on training history plots, model specification, and performance metrics in the classification report module. The deep learning models include One-Dimensional Convolution Neural Networks (1D CNN), Long-Term Short Memory (LSTM), and Gated Recurrent Unit (GRU). The news data are obtained from the Kaggle site, containing 41.726 rows of data. Based on the experiments with the text data that has been preprocessed in the form of vectors and the specific parameters before starting, the results show that GRU achieves the highest performance value in accuracy, recall, precision, and F1 score. Although GRU becomes the model with the smallest file size, it is the slowest model to generate predictions from text news data. It also has a higher potential to be an overfitted model due to parameters than a simple RNN.

Evaluating Airline Passengers’ Satisfaction during the COVID-19 Pandemic: A Case Study of AirAsia Services through Sentiment Analysis and Topic Modelling

Lee Jie Yu, Nor Hasliza Md Saad, Zhu Kun, Ghada ElSayad — Wed, 21 Aug 2024 00:00:00 +0000

AirAsia has emerged as a dominant force among prominent low-cost airlines in recent years. However, the COVID-19 pandemic outbreak has severely impacted airline services, including AirAsia. There is a strong need for airline services to monitor customer experience and satisfaction from online customer reviews on the website to keep pace with changing customer perceptions toward their service quality. A growing number of travelers choose to express their experiences and emotions on online customer review platforms, resulting in substantial online airline service evaluations. The research analyzes 796 online customer reviews from Skytrax, a well-known online airline review website. The information hidden in customer-generated reviews is analyzed with the text mining technique, including topic modeling and sentiment analysis. The research uses the Latent Dirichlet Allocation (LDA) model for topic analysis and the Valence Aware Dictionary for Sentiment Reasoning (VADER) model for sentiment analysis. The sentiment ratio for AirAsia’s online reviews is approximately 59% positive and 41% negative. Only four reviews are neutral. The findings indicate that the online review of AirAsia has a greater proportion of positive sentiments than negative sentiments. In addition, the topic modeling shows hidden topics with the top high-probability keywords concerned with interior and seat, baggage, online service, staff service, flight schedule, and refund. The research demonstrates using sentiment analysis and topic modeling on customer review data as a more thorough alternative to survey-based models for researching airline service. The research contributes to the methodological advancements in text mining analysis and expands the current knowledge of customer review data.

Determinant Factors of Logistics Firm Performance Mediated by Optilog Adoption Using a Mixed Method with an Explanatory Sequential Analysis

Melianie, Rindang Widuri, Indra Gamayanto, Arta Moro Sundjaja — Tue, 27 Aug 2024 00:00:00 +0000

Recently, trucking logistics service providers are focusing on adopting technology and Information Technology (IT) capability to improve performance. Hence, it is essential to investigate how the IT capability of the firm can enhance logistics performance and impact technology adoption on performance. Therefore, the research examines the determinant factors of logistics firm performance mediated by Optilog adoption. The research design is a mixed method with an explanatory sequential design using quantitative and qualitative data analysis. Quantitative data analysis uses Structural Equation Modeling (SEM), while qualitative analysis adopts in-depth interviews. The analysis population includes 41 Optilog users in the logistics firm in East Jakarta, with the experts adopting a census sampling method. For qualitative phase, the researchers also recruit four respondents holding supervisory positions or higher using Optilog for daily operations. The results show that logistics firm performance is influenced by service excellence and Optilog adoption. Additionally, perceived risk, traceability, and IT capability positively affect Optilog adoption, which mediates the effects on logistics firm performance. The results confirm the rejection of the hypothesis that Optilog adoption mediates IT capability on logistics firm performance. The thematic analysis further concludes that the system, implemented in March 2022, does not impact logistics firm performance significantly, according to the respondents. Consequently, IT departments depend on responsiveness to address users’ needs and issues since when the vendor develops the system, it faces significant constraints due to the development and maintenance costs.

Deciphering Digital Discourse: Detecting Cyberbullying Patterns in Filipino Tweets Using Machine Learning

January F. Naga, Rabby Q. Lavilles — Tue, 10 Sep 2024 00:00:00 +0000

The research addresses the escalating challenge of cyberbullying in the Philippines, a concern magnified by widespread social media use. A dataset of 146,661 tweets is analyzed using a pre-trained natural language processing model tailored to detect derogatory Filipino terms. The methodology is designed to preprocess data for clarity and analyze derogatory phrases, using the 23 key terms to indicate cyberbullying. Through quantitative analysis, specific patterns of derogatory term co-occurrence are uncovered. The research specifically focuses on Filipino digital discourse, uncovering patterns of derogatory language usage, which is unique to this context. Combining data mining and machine learning techniques, including Frequent Pattern (FP)-growth for pattern identification, cosine similarity for phrase correlation, and classification technique, the research achieves an accuracy rate of 97.91%. To assess the model’s reliability and precision, a 10-fold cross-validation is utilized. Moreover, by examining specific tweets, the analysis highlights the alignment between automated classifications and human judgment. The co-occurrence of derogatory terms, identified through methods like FP-growth and cosine similarity, reveals underlying cyberbullying narratives that are not immediately obvious. This approach validates the high accuracy of the models and emphasizes the importance of a comprehensive framework for detecting cyberbullying in a linguistically and culturally specific context. The findings substantiate the effectiveness of the targeted approach, providing essential insights for developing cyberbullying prevention strategies. Furthermore, the research enriches the literature on digital discourse analysis and online harassment prevention by addressing cyberbullying patterns and behaviors. Importantly, the research offers valuable guidance for policymakers in crafting more effective online safety measures in the Philippines.

Indonesian-English Textual Similarity Detection Using Universal Sentence Encoder (USE) and Facebook AI Similarity Search (FAISS)

Tue, 10 Sep 2024 00:00:00 +0000

The tremendous development in Natural Language Processing (NLP) has enabled the detection of bilingual and multilingual textual similarity. One of the main challenges of the Textual Similarity Detection (TSD) system lies in learning effective text representation. The research focuses on identifying similar texts between Indonesian and English across a broad range of semantic similarity spectrums. The primary challenge is generating English and Indonesian dense vector representation, a.k.a. embeddings that share a single vector space. Through trial and error, the research proposes using the Universal Sentence Encoder (USE) model to construct bilingual embeddings and FAISS to index the bilingual dataset. The comparison between query vectors and index vectors is done using two approaches: the heuristic comparison with Euclidian distance and a clustering algorithm, Approximate Nearest Neighbors (ANN). The system is tested with four different semantic granularities, two text granularities, and evaluation metrics with a cutoff value of k={2,10}. Four semantic granularities used are highly similar or near duplicate, Semantic Entailment (SE), Topically Related (TR), and Out of Topic (OOT), while the text granularities take on the sentence and paragraph levels. The experimental results demonstrate that the proposed system successfully ranks similar texts in different languages within the top ten. It has been proven by the highest F1@2 score of 0.96 for the near duplicate category on the sentence level. Unlike the near-duplicate category, the highest F1 scores of 0.77 and 0.89 are shown by the SE and TR categories, respectively. The experiment results also show a high correlation between text and semantic granularity.

Data Monetization Service Development Using Iterative Lifecycle Framework, Quality Assurance, and Open Web Application Security Project: A Case Study of a Utility Company in Indonesia

Thu, 19 Sep 2024 00:00:00 +0000

The research aims to provide Data Monetization (DM) services for an Indonesian utility company as a pilot to generate additional revenue beyond the primary operation. The service is built using an iterative development lifecycle framework and evaluated based on five Quality Goals (QGs), including application and security testing activities. The framework includes methods for processing and modeling electricity usage data, testing application quality, checking infrastructure quality, and ensuring access security for front-end and back-end applications using the Open Web Application Security Project (OWASP). For data modeling, Support Vector Regression (SVR) is used, and it outperforms Polynomial Regression (PR) and Multi-Layer Perception (MLP) Neural Networks. Furthermore, QG shows strong performance with an Relative Root Mean Squared Error (RRMSE) value < 10%, high forecasting ability with Mean Average Probability Error (MAPE) < 10%, and a near-zero average error rate (Mean Squared Error (MSE)) square using minimal data from four months. The services go through functional and integration test to ensure product quality and application performance, which results in a minimum of 95% service response in throughput, 0.128 seconds for processing 2,000 requests, stability at 300–500 in one second per hour, and 7–21 seconds during peak hours. Additionally, the service passes nine penetration tests and ten vulnerability assessments using the OWASP top 10:2021 category. Based on the comprehensive testing and evaluation results, both the application and the service are considered ready and secured for deployment.

Fuzzy-Based Decision Support Model for Assessing Green Building Performance

Muhamad Akbar Bin Widayat, Ditdit Nugeraha Utama — Thu, 19 Sep 2024 00:00:00 +0000

Global warming is currently a major environmental issue that is capable of causing unpredictable climate changes. The phenomenon is due to the accumulation of gases and carbon dioxide in the earth’s atmosphere, partly attributed to building operation and construction. The Green Building Rating System (GBRS) is developed to assess and measure the level of green building practices to address this problem. The assessments have typically been conducted using conventional methods that require parameters to meet specific criteria. However, certain parameter values cannot be calculated using objective methods, such as bias, time series, and distance values. The existence of these challenges leads to the development and integration of the Decision Support Model (DSM) into the GBRS in the research. The DSM uses a mathematical model, Tsukamoto Fuzzy Inference System (FIS), and conventional methods to handle the parameter values. Moreover, data related to the parameters are collected and analyzed quantitatively. As a result, the DSM-GBRS model is successfully implemented with two findings. First, there are 83 parameters, related to policy, retrofit, construction, and utilization aspects based on Peraturan Menteri Pekerjaan Umum dan Perumahan Rakyat Nomor 21 Tahun 2021. Second, the model provides precise decision values by splitting the treatment into four types: conventional, Fuzzy logic, slope, and Euclidean distance to ensure a comprehensive assessment of green building performance.

Smart Aquaculture Design for Vannamei Shrimp Farming Based on Quality Function Development

Budi Setiawan, Nico Surantha — Mon, 30 Sep 2024 00:00:00 +0000

In the fishery industry, Indonesia’s large water area has the potential for developing and cultivating fisheries such as vannamei shrimp. For this reason, aquaculture, particularly vannamei shrimp farming, can play a crucial role in Indonesia’s economy and food supply. However, challenges such as fluctuating water quality, disease outbreaks, turbidity levels, and irregular shrimp feeding schedules in ponds can affect the productivity and sustainability of shrimp farming. The smart aquaculture system integrates technologies, such as IoT-based sensors, automated feeding mechanisms, and real-time water quality monitoring to optimize the farming process. The research proposes a smart aquaculture design for vannamei shrimp farming based on the Quality Function Development (QFD) method. It starts by creating questionnaires to identify stakeholders’ level of interest. The questionnaire results are used as a reference for system redesign using the QFD method to improve the quality and quantity of shrimp harvest, cultivating effectively and efficiently and helping and facilitating the supervision of pond managers on pond water quality, feeding, and feed availability. The result highlights the application of QFD in creating a tailored, technology-driven solution that supports better decision-making, resource optimization, and improved shrimp health. The system reduces human error, enhances farm management, and promotes higher yields by providing real-time data and automation. The evaluation results show that the proposed design can achieve high stakeholder satisfaction. It also achieves better scores compared to the other two competitor’s designs.

Modeling Emotion Recognition System from Facial Images Using Convolutional Neural Networks

Jasen Wanardi Kusno, Andry Chowanda — Fri, 04 Oct 2024 00:00:00 +0000

Emotion classification is the process of identifying human emotions. Implementing technology to help people with emotional classification is considered a relatively popular research field. Until now, most of the work has been done to automate the recognition of facial cues (e.g., expressions) from several modalities (e.g., image, video, audio, and text). Deep learning architecture such as Convolutional Neural Networks (CNN) demonstrates promising results for emotion recognition. The research aims to build a CNN model while improving accuracy and performance. Two models are proposed in the research with some hyperparameter tuning followed by two datasets and other existing architecture that will be used and compared with the proposed architecture. The two datasets used are Facial Expression Recognition 2013 (FER2013) and Extended Cohn-Kanade (CK+), both of which are commonly used datasets in FER. In addition, the proposed model is compared with the previous model using the same setting and dataset. The result shows that the proposed models with the CK+ dataset gain higher accuracy, while some models with the FER2013 dataset have lower accuracy compared to previous research. The model trained with the FER2013 dataset has lower accuracy because of overfitting. Meanwhile, the model trained with CK+ has no overfitting problem. The research mainly explores the CNN model due to limited resources and time.

Editorial Page and Table of Content

Fergyanto E. Gunawan — Wed, 21 Aug 2024 00:00:00 +0000