UPLB ICS Peak One

This paper introduces a specialized degree planner web application designed for undergraduate students at the Uni- versity of the Philippines Los Ba˜ nos (UPLB). The planner aims to help students overcome challenges in following their prescribed curriculum by offering features such as prerequisite identification and unit requirement verification. It utilizes UPLB’s course and curriculum database to ensure accuracy and compliance with university rules. A user evaluation with UPLB undergraduates indicated above-average usability, though feedback suggested improvements in instructions and mobile-friendliness. Future work includes refining usability and exploring the planner’s impact on student success metrics.

The technological advancement of Generative Ad- versarial Networks (GANs) has allowed the creation of synthetic images, posing a threat for digital disinformation and media fabrication. Due to this, numerous methods have been pro- posed to counter GAN-generated media. However, most methods employ deep learning and the use of Convolutional Neural Networks (CNNs), which can be computationally expensive to train. This study proposes frequency analysis through Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT) as distinct pre-processing methods for GAN image classification. Additionally, this study uses a Support Vector Machine (SVM) model to classify fake from real images. To address the limitations of using faces as the primary object class, this study investigates the generalizability of GAN traces in the frequency domain across various object classes using the ProGAN dataset. The study found that using DCT as a pre-processing method provides the most significant performance among the proposed methods, with an accuracy of 97.08%.

This project aims to create a transparent platform that displays the demand for elective subjects offered by various colleges at the University of the Philippines - Los Baños. By providing real-time data on student interest and enrollment trends, this system will enable better course planning for students, allowing them to make informed decisions about their elective choices.

This study explores the integration of machine learning into e-commerce platforms for automated product legitimacy detection, focusing on Shopee’s Philippine site. By developing a Chrome browser extension that applies TF-IDF for text vectorization and logistic regression for classification, the system analyzes user reviews to determine the legitimacy of products. The study aims to protect consumers from counterfeit products while fostering an authentic marketplace for legitimate sellers. Data was collected from Shopee reviews, which were used to train and evaluate the model. The model achieved an accuracy of 84.5%, along with high precision, recall, and F1- scores. In terms of usability, a System Usability Scale (SUS) score of 76.72 confirmed the extension as user-friendly and effective, providing a reliable solution to improve trust and transparency in e-commerce.

This paper presents the design and implementation of an Interoperable RESTful API-Driven Database System. Functioning as a standardized repository, this database system aims to consolidate comprehensive information on the microbial diversity of bat and bat guano from CALABARZON caves. Leveraging RESTful Web Services, the implementation ensures a fluid exchange of information across diverse research systems, prioritizing both data consistency and security. Through a series of tests conducted, including unit testing, speed testing, and interoperability testing, the outcomes attest to the development of a resilient, secure, and highly efficient system.