Max Angelo Dapitilla Perin

Mr. Max Angelo Dapitilla Perin

Bohol Island State University, City of Tagbilaran, Philippines
+5
orcid
Bohol Island State University | ᜉᜋᜈ᜔ᜆᜐᜅ᜔ ᜉᜋ᜔ᜉᜋᜑᜎᜀᜈ᜔ ᜅ᜔ ᜁᜐ᜔ᜎ ...

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 January 2022
Verified
Conference Paper
Open access
Max Angelo Dapitilla PerinMax Angelo Dapitilla Perin
Max Angelo Dapitilla Perin
LFLarmie Santos Feliscuzo
Larmie Santos Feliscuzo
CRCherry Lyn Cando Sta. Romana
Cherry Lyn Cando Sta. Romana
A language dies when the population that speaks it fails to pass it to their future generations. This study aims to address such problems for the Eskayas and Boholanos. The proposed EskayApp is an OCR Transliteration e-Learning Android Application that will serve as a tool to learn Eskaya. The app uses the k-Nearest Neighbors Algorithm (k-NN) for its business logic layer or backend and Supervised Machine Learning to create the application programming interface (API) for the said application. The Android application accesses the camera to capture an image of an Eskaya character and returns the Latin transliteration and the equivalent Latin character. The application got a 4.524-star rating through Google Play Store. After conducting testing, the machine learning model accuracy was 89.93% based on a 2x2 confusion matrix. A usability test was also conducted with members of the tribe as respondents. The usability review has 3 measurement items in which users can respond from 1 (Very Poor) to 5 (Excellent). The average for all the measurement items gathered from the users is 4.3-star and all the respondents want to use the e-Learning application again.

11 April 2024
Verified
Conference Paper
Open access
Darrel CardañaDarrel Cardaña
Darrel Cardaña
Max Angelo Dapitilla PerinMax Angelo Dapitilla Perin
Max Angelo Dapitilla Perin
STSheila Garcia Tabuno
Sheila Garcia Tabuno
EOElizabeth Ancog Orapa
Elizabeth Ancog Orapa
MDMae Suaybaguio Dagupan
Mae Suaybaguio Dagupan
JLJoan Mae Garcia Lagumbay
Joan Mae Garcia Lagumbay
Stenography is a technology that faithfully turns physically audible speech into a precise and permanent written record at the real-time moment in which the recorded speech is uttered. As a human can no longer interpret massive amounts of data thoroughly and accurately, the application of machine learning aids in advancing artificial intelligence. Gregg Shorthand in stenography is the art of writing in shorthand or using specialized symbols and abbreviations. The project aims to investigate the many uses of image processing in steganography. The research utilizes a descriptive-developmental research design. Specifically, supervised machine learning methods, particularly k-Nearest Neighbors (kNN), detect and convert handwritten Gregg shorthand symbols into readable Latin letters. The dataset consists of forty-eight images. The result suggested that k-Nearest Neighbors (kNN) works well in identifying images with a precision performance of 86% and overall performance using the ISO 25010 of 4.17 interpreted as the system is very usable.

22 March 2024
Verified
Conference Paper
Closed access
Max Angelo Dapitilla PerinMax Angelo Dapitilla Perin
Max Angelo Dapitilla Perin
DCDarrel Abuyabor Cardaña
Darrel Abuyabor Cardaña
CGCecilia Tabel Gumanoy
Cecilia Tabel Gumanoy
Writing systems have played an essential role in human communication, evolving uniquely within various cultures and societies. One such writing system is Baybayin, an ancient script for writing Philippine languages that has recently rekindled interest. This study dedicates itself to developing a Baybayin Character Recognition and Transliteration mobile application that utilizes the k-Nearest Neighbors (k-NN) Algorithm. The primary objective of this mobile application is to enhance the accessibility of Baybayin by enabling users to recognize and transliterate images of fundamental Baybayin characters. Utilizing the k-NN algorithm within the API enhances accuracy through machine learning techniques. Before we make the app available to users, we will test images of primary Baybayin characters captured on mobile phones for accuracy. After assessing the machine learning model API using a 2x2 confusion matrix, it achieved an accuracy rate of 95.47%. The researchers recommend implementing the developed system in schools that teach Baybayin. Further, future researchers may conduct studies using empirical evidence and impact assessment of utilizing the developed system correlated with students' efficiency and performance.
1

3 September 2025 in Journal of Technology-Assisted Learning
Verified
Journal Article
Open access
PDF file
Max Angelo Dapitilla PerinMax Angelo Dapitilla Perin
Max Angelo Dapitilla Perin
The increasing integration of digital technologies into academic administration has significantly improved operational efficiency, particularly in attendance monitoring. Traditional methods, such as manual sign-ins and biometric scanners, often have limitations, including time inefficiency, susceptibility to human error, and administrative burden. To address these issues, this study evaluated the usability and performance of the BISU-Bilar QR Code Generator and Contact Tracing System, which was repurposed for monitoring student attendance at Bohol Island State University–Bilar Campus. A mixed-methods research design was employed, combining quantitative data from structured surveys and qualitative insights from open-ended responses. The evaluation was guided by the ISO 25010 standard, focusing on key usability dimensions such as functional suitability, efficiency, reliability, and user satisfaction. Results from 106 respondents indicate that the system is generally functional, easy to learn, and effective in facilitating task completion. The highest-rated usability aspect was device compatibility, while system stability and security received the lowest scores, suggesting areas for technical improvement. Statistical analyses, including Principal Component Analysis (PCA) and Ordinary Least Squares (OLS) regression, revealed that system functionality and device compatibility significantly influence perceived efficiency. In contrast, system crashes and data security concerns negatively impact user experience. Visualized through a word cloud, qualitative feedback highlighted recurring themes such as the demand for real-time notifications, enhanced user interface design, and improved technical support. The study concludes that while the QR code-based system offers a practical and scalable solution for attendance tracking, targeted enhancements in system stability, cross-platform performance, and data protection are necessary to ensure reliability, boost user confidence, and promote long-term adoption within educational institutions.

30 July 2025 in International Journal of Recent Technology and Engineering (IJRTE)
Verified
Journal Article
Open access
Max Angelo Dapitilla PerinMax Angelo Dapitilla Perin
Max Angelo Dapitilla Perin
LFLarmie S. Feliscuzo
Larmie S. Feliscuzo
CAChris Jordan G. Aliac
Chris Jordan G. Aliac
NCNelia Q. Catayas
Nelia Q. Catayas
This study presents the development and evaluation of the Local Irrigation System Assessment (LISA), a web-based platform designed to monitor farm water availability and support informed irrigation management. The core objective of the research is to address the lack of accessible, real-time irrigation monitoring tools for rural farming communities, particularly in low-bandwidth environments. LISA was developed using open-source web technologies, with a focus on usability, responsiveness, and accessibility. The system features a dashboard for monitoring water availability, user-specific data visualization, and notification alerts to assist farmers and agricultural stakeholders make timely irrigation decisions. The development process followed a user-centred and iterative design approach to ensure simplicity for users with varying levels of digital literacy. To evaluate the platform's technical quality, the study employed Google Lighthouse, an automated auditing tool for web performance. The website was assessed across four key metrics: Performance, Accessibility, Best Practices, and SEO (Search Engine Optimization). Evaluation results demonstrated high scores in Accessibility and SEO, indicating strong compliance with global web standards and discoverability. Best Practices scores confirmed adherence to modern development protocols, while Performance scores identified optimization opportunities, particularly in image compression and script loading. This research contributes a practical framework for building web platforms that meet critical technical standards in rural and agricultural contexts. The study highlights how digital platforms can be optimized for broader adoption and greater impact by integrating usability with performance evaluation tools. The LISA platform exemplifies how data-driven tools can be leveraged to empower local farmers and improve the effectiveness of irrigation systems through timely, accessible, and actionable insights.

15 January 2025 in Technologique: A Global Journal on Technological Developments and Scientific Innovations
Verified
Journal Article
Closed access
CTCebu Institute of Technology-University
Cebu Institute of Technology-University
Max Angelo Dapitilla PerinMax Angelo Dapitilla Perin
Max Angelo Dapitilla Perin
LFLarmie Feliscuzo
Larmie Feliscuzo
CAChris Jordan Aliac
Chris Jordan Aliac
NCNelia Catayas
Nelia Catayas
BCBohol Island State University-Bilar Campus
Bohol Island State University-Bilar Campus
Irrigators' Associations (IAs) are critical in managing water resources and influencing agricultural productivity and rural sustainability. Predicting IA membership is complex due to socio-economic, organizational, and environmental factors. This study applies machine learning techniques—Decision Trees and Random Forests— to model IA membership and identify the key predictive variables. Using a dataset of 234 IA records, the models were evaluated based on Mean Absolute Error (MAE), with the Random Forest model achieving a lower MAE of 19.49, compared to 24.92 for the Decision Tree. Key predictors include the number of farm beneficiaries, service area, years of operation, and leadership structure. The results demonstrate the potential of machine learning in supporting data-driven planning for IA engagement, offering valuable insights that can enhance resource allocation and inform policy development in the agricultural sector, ultimately contributing to more efficient water management and improved rural livelihoods.