UPLB ICS Peak One

Several college students experience inconvenience when accessing printing services. This study proposes ”UPrinter,” a peer-to-peer (P2P) web-based system that connects students in need with students who are willing to offer their printing service. The system accommodates features such as print requests, service offers, direct messaging, and feedback gathering. The usability and user perceptions of the platform were measured with the System Usability Scale (SUS) and qualitative feedback from 22 University of the Philippines Los Ba˜ nos students. This mixed- methods research explored the usability, convenience, and effect of the platform on productivity among students. The results emphasize the potential of P2P models to provide easy-to-use and adaptable printing alternatives that promote resource sharing within the university.

UPLB students often rely on multiple applications or platforms to manage their academic tasks, leading to frag- mented organization and inefficiencies. IskoTrack addresses this problem by providing a centralized, offline-capable mobile appli- cation designed to help students manage their course schedules, deadlines, and academic activities—reducing cognitive load and enhancing productivity in navigating academic responsibilities. Developed using Flutter, the study aimed to provide a cross- platform mobile application that streamlines academic tracking and improves availability for students regardless of internet connectivity. The mobile application was tested and evaluated by 26 respondents composed of undergraduate UPLB students using both Android and iOS devices. The evaluation yielded a System Usability Scale (SUS) score of 75.96%, which translates to good and acceptable usability. alongside a Questionnaire for User Interface Satisfaction (QUIS) mean score of 8.08 out of 10, reflecting a high level of user satisfaction with the application’s interface.

This study addresses data management inefficien- cies within local government, particularly focusing on the Santa Cruz Traffic Management Office (SCTMO) in Laguna, Philip- pines. To address these issues, The SCTVMS (Santa Cruz Traffic Violations Management System) was developed, a web application built to centralize data management operations. The system implements role-based access control, Create, Read, Update, Delete (CRUD) functionalities, data visualization, and reporting capabilities tailored to the office’s needs. System eval- uation conducted with 25 municipal employees using the System Usability Scale (SUS) yielded an excellent average score of 84.8. The implementation demonstrates that meaningful digital trans- formation at the local government level can significantly enhance operational efficiency and public service delivery, providing a model for other municipalities facing similar challenges.

The integration of loanwords from English into Tagalog presents unique challenges in natural language pro- cessing, particularly in morphological analysis and stemming tasks. This study introduces a rule-based stemmer for English words with Tagalog morphology, informed by the morphological structure tabulation of Sulit et al. (2024). The stemmer leverages insights into common affixation, partial duplication, and com- pounding structures of English loanwords in Tagalog, enabling robust segmentation into stems. The stemmer was evaluated on an annotated dataset, with accuracy of 91.1960%. This work contributes to low-resource NLP by providing a reusable rule- based framework, a linguistically-informed dataset, and insights into the computational modeling of code-mixed languages. The stemmer lays foundational groundwork for further research in Taglish and other code-mixed language processing tasks.

Manual coin counting can be time-consuming, es- pecially for individuals who frequently deal with coins. Coin counting machines, while effective, are often costly for these users. This study presents a mobile application designed to count Philippine Peso coins using YOLOv5. A comparative analysis of three object detection models, YOLOv5, YOLOv8, and RT- DETR, was conducted to identify the most suitable model. YOLOv5 was chosen primarily for its competitive results, small model size, and fast inference speed, making it the best fit for mobile deployment. The application was tested in various real-world scenarios, achieving an overall accuracy of 94.2%. It also garnered a mean SUS score of 79, indicating positive user satisfaction. Overall, this mobile application offers a practical, cost-effective alternative to traditional coin counting methods, with strong real-world performance and user acceptance.