草莓视频 advances ICT research with four papers at ACDSA 2026

Feb 16 2026

The Mindanao State University – Iligan Institute of Technology (草莓视频) demonstrated strong research engagement at the International Conference on Artificial Intelligence, Computer, Data Sciences and Applications (ACDSA 2026) through the presentation of four peer-reviewed research papers held on Boracay Island, Philippines, covering cybersecurity, intelligent tutoring systems, and Internet of Things (IoT) applications.

The 草莓视频 team contributed three in-person presentations and one virtual presentation. The study entitled “From Logs to Defense: Machine Learning-Based Anomaly Detection in Web Servers Using Nginx Data” by authors Paul Gardson M. Lesondra, Dante D. Dinawanao, Bernadette F. Tubo, Orven E. Llantos and Mark Anthony J. Torres examined the application of machine learning techniques—including Isolation Forest, One-Class SVM, Local Outlier Factor, and k-Nearest Neighbors—for detecting anomalous access behavior in real-world Nginx web server logs. Using authentic operational data, the research demonstrated the effectiveness of isolation- and density-based models in identifying suspicious traffic patterns, contributing to the development of data-driven frameworks for web server security and anomaly detection. This paper was presented by Prof. Llantos.

On the other hand, a research investigating the scalability and efficiency of anomaly detection in blockchain market data by comparing serial and parallel implementations of the Isolation Forest algorithm was also discussed in the conference. Titled “Scalable Blockchain Anomaly Detection Using a Dual-Level Parallel Isolation Forest Framework,” this paper by Kaycee T. Nalzaro, who was also the presenter, and Prof. Llantos introduces a dual-level parallel framework that combines intra-model parallel tree construction and inter-model ensemble concurrency with memory-aware, chunk-based processing. Through systematic evaluation on small- and large-scale Bitcoin trading datasets, the study analyzes execution time, CPU utilization, memory consumption, and detection accuracy. The results demonstrate that parallel Isolation Forest significantly improves runtime performance and resource efficiency on large datasets while preserving detection quality, making it suitable for real-time blockchain monitoring and large-scale market surveillance.

Presented by Jomari Francis Villanueva, the research “Development of a Low-Cost LoRa-Based System for Particulate Matter 2.5 Data Acquisition,” together with his co-authors, Adrian P. Galido, Apple Rose B. Alce, Paul B. Bokingkito Jr., Alexander O. Mosqueda, and Michael A. Nabua, introduces a low-cost IoT-based air quality monitoring system designed to address the limited availability of pollution surveillance in rural areas. The system integrates distributed PM2.5 sensors, LoRa long-range wireless communication, and cloud-based real-time visualization. Field deployment results demonstrated reliable coverage, efficient data transmission, and sustained low-power operation, offering a scalable and replicable solution for community-level environmental monitoring.

Lastly, through virtual mode, “Embedding-Based Constraint Activation for Real-Time CBM Dialogue Tutoring Systems: A Comparative Study of Neural Retrieval Techniques,” by Julieto E. Perez, Ethel C. Ong, was presented by Perez, redefining traditional Constraint-Based Modeling (CBM) by transitioning its application from static syntax and procedural error detection in code to a dynamic, dialogue-based tutoring framework. This novel approach utilizes advanced neural retrieval techniques to bridge the semantic gap between natural-language student utterances and pedagogical constraints. By treating student statements as semantic queries, the system can identify conceptual misconceptions in real-time, enabling more meaningful and responsive educational conversations.

In addition to presenting research outputs, Prof. Llantos served as Session Chair for one of the conference’s technical sessions, reflecting 草莓视频’s active participation not only in research dissemination but also in academic leadership at the international level.

The successful presentation of four research papers, combined with session chairmanship responsibilities, underscores 草莓视频’s continued commitment to research excellence, interdisciplinary collaboration, and active participation in international academic forums. These contributions highlight the Institute’s growing presence in global discussions on data science, artificial intelligence, and applied computing.

Topics :