IJCONSIST JOURNALS

Evaluating CNN-LSTM for Indonesian Hoax News Classification: A Comparative Analysis with Existing Models

2026-10-12T02:37:07+00:00

Abstract— The spread of Indonesian-language hoax news through digital platforms continues to increase along with the massive use of social media and online news portals. This condition poses a serious challenge to providing accurate and reliable information to the public, necessitating an effective AI-based system for identifying hoax news. This study aims to implement the Convolutional Neural Network–Long Short-Term Memory (CNN-LSTM) algorithm in an Indonesian-language hoax news identification system. The CNN-LSTM method was chosen because it combines the advantages of CNNs in extracting local text features and LSTMs in modeling contextual word-order dependencies. The dataset used consists of a collection of Indonesian-language news articles that have been labeled as hoax or non-hoax, after text preprocessing stages including cleaning, tokenization, stopword removal, and padding. Model performance was evaluated using the Accuracy, Precision, Recall, and F1-Score metrics. The test results show that the CNN-LSTM model can classify hoax and non-hoax news with good accuracy and consistency, demonstrating the effectiveness of the hybrid approach for hoax detection. Based on these results, it can be concluded that the implementation of CNN-LSTM has the potential to be a reliable solution to support the Indonesian-language hoax news identification system.

Keywords: hoax news, CNN-LSTM, text classification, deep learning, Bahasa Indonesia

Implementation of OSPF and Static Routing Using Python Scripts in Cisco Packet Tracer Network Simulation

2026-10-17T23:22:17+00:00

The increasing complexity of computer networks demands a faster and more accurate process of configuring network devices. Manually configuring routers is time-consuming and prone to configuration inconsistencies, requiring a more efficient approach. This study aims to implement static routing and dynamic routing configurations based on the OSPF (Open Shortest Path First) protocol using the Python programming language in a simulation of the Cisco Packet Tracer network. The network topology consists of three routers, three switches, and three PCs with predefined IP addressing. Python scripts are designed to generate router interface configurations, static routing, and OSPF based on topological data, as well as store the results into .txt files. The test is performed using ping commands, show ip route, and show ip ospf neighbor to verify inter-network connectivity. The test results showed that both routing methods successfully connected three different subnets with 0% packet loss under stable network conditions. Python-based scripts have been shown to speed up the configuration preparation process as well as produce more consistent and structured configurations compared to manual methods.

Implementation Of Face Distance, Usage Duration, And Smart Break Monitoring System Using Python-Based Haar Cascade On Digital Device Usage

2026-10-19T23:25:10+00:00

The prolonged use of digital devices with improper viewing distance can lead to eye strain and visual discomfort. This study aims to develop a real-time monitoring system for digital device usage based on digital image processing. The system is implemented using the Python programming language and the OpenCV library, with the Haar Cascade Classifier method used to detect the user's face through a webcam. Distance estimation is performed using single-camera distance estimation based on facial anthropometric data and the width of the detected face in the image, and is stabilized using smoothing techniques. In addition, the system utilizes user age input to determine the recommended viewing distance range, calculates device usage duration, and evaluates the level of usage risk based on distance, age, and usage time. The system provides warnings in the form of visual alerts, beep sounds, break reminder audio, and a Smart Break feature that requires the user to rest for 20 seconds after a certain usage duration is reached. The implementation results show that the system is capable of detecting faces in real time, estimating distance with stable performance, and providing adaptive and responsive usage notifications. Therefore, the proposed system can be used as a simple and effective eye health monitoring tool without requiring additional hardware.

Development of a Sensitive-Data Monitoring System for Witness and Victim Protection: Role-Based Security Integration and Usability Evaluation

2026-10-19T23:24:07+00:00

Data on applications, case reviews, and protection services for witnesses and victims at the Medan Representative Office of the Witness and Victim Protection Agency (LPSK) had previously been managed through a legacy application built without a modern framework, which constrained data-handling efficiency, feature coverage, and the security controls required for information tied directly to personal safety. Existing literature widely discusses public-service monitoring in general terms but rarely demonstrates concrete, tested technical safeguards for highly sensitive data such as the identities of protected individuals. This study develops a web-based monitoring platform on the Laravel framework that consolidates the application, review, service, and financial workflows into a single system, protected by layered security consisting of Role-Based Access Control (RBAC), role-based data masking, Bcrypt password hashing, CSRF protection, rate limiting, UUID-based routing, and audit trails. A Research and Development (R&D) approach following the 4D model (Define, Design, Develop, Disseminate) was used. Black-box functional testing across 35 scenarios and security testing across 10 scenarios each achieved a 100% success rate, while a User Acceptance Test with 5 respondents at LPSK Medan produced a mean System Usability Scale (SUS) score of 93.6 (Grade A, Excellent). The principal contribution of this work is the integrated combination of RBAC, role-based data masking, and audit trails, implemented and empirically validated within a highly sensitive government information system. System scalability was not verified through load testing, and the security evaluation did not include independent penetration testing, both of which remain limitations and directions for future work.

Sentiment Analysis of Google Gemini Reviews Using K-Nearest Neighbor and Particle Swarm Optimization

2026-10-19T23:23:46+00:00

Google Gemini, Google's generative AI assistant, receives a large volume of user reviews on the Google Play Store daily, making manual reading of user opinions impractical. This study analyzes the sentiment of Indonesian-language Google Gemini reviews using K-Nearest Neighbor (KNN) as the baseline classifier, then optimizes the k parameter through Particle Swarm Optimization (PSO) to examine its effect on classification performance. A total of 5,000 reviews were collected via web scraping using the Google Play Scraper library between 1 June and 1 September 2026, followed by preprocessing (duplicate removal, cleaning, case folding, normalization, tokenization, and stopword removal), leaving 2,858 labeled reviews, of which 2,000 were used for modeling. Negative, Neutral, and Positive labels were assigned automatically from each reviewer's star rating. Text was represented using CountVectorizer, producing 2,282 features, split into 1,600 training and 400 testing samples via stratified sampling. The baseline KNN model with k=5 achieved 79.50% accuracy, while PSO selected k=4 with a best cost of 0.188125, yielding a KNN-PSO accuracy of 80.75%, a 1.25 percentage-point increase. This improvement was accompanied by a rise in macro F1-score from 0.441 to 0.458 and weighted F1-score from 0.765 to 0.780, mainly driven by the Negative and Positive classes, while the Neutral class remained undetected by both models (F1=0). These findings indicate that PSO optimization yields a consistent but limited improvement, and does not resolve the underlying difficulty of identifying

Implementation of a Telegram Bot for Router Monitoring Using Python and the RouterOS API

2026-10-19T23:24:53+00:00

Network monitoring plays an important role in maintaining the stability and availability of computer network services. Manual monitoring approaches often cause delays in detecting network faults and responding to disruptions. This study implements a MikroTik router monitoring system that integrates Python, the RouterOS API, a Telegram Bot, and a Streamlit dashboard to support real-time fault detection and administrator notification. The system periodically checks router connectivity status and automatically detects when a router transitions between online and offline states. Upon detecting a status change, the system delivers immediate notifications to administrators through Telegram Bot and records the event in a log file for diagnostic purposes. Router metrics including CPU usage, memory availability, uptime, and interface conditions are displayed through a web-based interactive dashboard, providing administrators with a visual interface for monitoring network health. System testing was conducted by simulating router online and offline conditions. Results show that the system successfully detects router status changes, maintains diagnostic logs, and delivers real-time notifications via Telegram and a web-based monitoring interface. This integration of automated fault detection, mobile notification, and web-based visualization offers an effective approach to improving network monitoring responsiveness.

AI-Based Wi-Fi Traffic Profiling and Anomaly Detection System

2026-10-19T23:24:33+00:00

Abstract— Network traffic monitoring plays an important role in maintaining the availability, reliability, and performance of modern network infrastructures. The increasing volume and complexity of network traffic require monitoring solutions that provide real-time visibility, traffic profiling, anomaly detection, and comprehensive usage analysis. This study proposes an AI-based Wi-Fi network traffic monitoring and analysis system that integrates ntopng for traffic collection, Python-based analytical modules, Prometheus for metrics aggregation, Grafana for visualization, K-Means Clustering for traffic profiling, and Isolation Forest for anomaly detection. The system was implemented in a wireless network environment using real network traffic data collected through the ntopng REST API. The collected traffic data were analyzed to identify traffic patterns, assess traffic severity, categorize network usage activities, and generate monitoring metrics for real-time visualization. The results demonstrate that the proposed system successfully performs traffic clustering, anomaly detection, usage categorization, dashboard-based monitoring, and automated alert integration. Furthermore, the generated analytical information provides insights into network utilization patterns that may support future Quality of Service (QoS) implementation, bandwidth allocation strategies, and intelligent network management. The integration of machine learning techniques with modern monitoring platforms contributes to more effective and centralized network monitoring.