UPLB ICS Peak One

Liver cancer research over the years has raised the necessity of shared research data among institutions. However, not all institutions designate the same term to refer to an object. This problem makes knowledge sharing difficult as groups must identify which terms mean which within an institution. This study created an application to help researchers make datasets by developing a compiled ontology system with con- trolled vocabulary generation using terms within the compiled ontologies. Researchers with expertise in Medicine or Bioinfor- matics and professionals familiar with the field assessed the web application’s usability using the System Usability Scale (SUS). Testing by users revealed a mean score of 81.2, thus showing excellent usability and ease of use in displaying various terms and efficiency in controlled vocabulary generation in data.

This paper addresses communication challenges faced by the deaf and hard of hearing community in the Philippines, particularly during emergencies. This paper presents an Android-based text translator for Filipino Sign Language (FSL) emergency signs, aiding communication with emergency responders. Using a dataset of FSL signs augmented with image processing techniques, a MobileNetV2, a convolutional neural network model, underwent transfer learning technique and was trained to recognize and translate signs into text. The mobile application, developed with Flutter, integrates this model, enabling real-time translation. Accuracy score, precision score, recall score, F1 Measure and System Usability Scale were used to evaluate the produced model and mobile application.

Variants are differences between a reference genome and a sample genome that can manifest as Single Nucleotide Polymorphisms, Insertions, Deletions, and Structural Variants. These Variants are important to study as they are mutations that can cause changes to an organism. Access to variant calling tools is usually gated behind expensive services, and while some are available as open source alternatives, they frequently have a steep learning curve. This study produced a web application that includes features for variant calling using a simple workflow whose output is in the form of VCF files that can be subsequently visualized via an integrated Genome Browser. The web applica- tion was created using Vue, Django, Docker, Conda, Snakemake, and various Bioinformatics software to allow for direct upload of VCF Files, as well as, implementation of a built-in variant calling workflow. The IGV Genome Browser was utilized and integrated seamlessly to be able to display the variants along with their reference genomes. The web application was evaluated by students in BS Biology and BS Computer Science with experience in Bioinformatics, along with professionals who are familiar with the field through the System Usability Scale (SUS). This resulted in an overall score of 77.65 which puts it above average in terms of Usability.

Creating final exam schedules is complex due to multiple constraints and limited resources. This study explores the application of satisfiability solvers in the automation of exam schedule generation. Leveraging modern solvers, a framework that can meet hard constraints (e.g., room availability) and soft constraints (e.g., minimal conflicting student schedules) can be made. Findings show that satisfiability solvers effectively create schedules with distinct course-to-room allocations while minimizing student conflicts.

Visualizing the changes to courses and retrieval of said changes in UPLB is difficult due to the reliance on paper proposals only and the absence of an electronic document management system for storing and managing digital versions of these information. Additionally, the current course viewer of AMIS, UPLB’s academic management system, only shows basic information about the course but not their relationship to other courses or the history of changes they have undergone. As such, this study aimed to enhance the course viewer of AMIS and provide a digital document management system for course proposals by developing an EDMS for managing, retrieving, and viewing courses and their proposals. This study also created a module for converting paper proposals to their digital form. 11 participants, consisting of UPLB faculty members and administrative staff, were tasked to evaluate the application. The application achieved a score of 62.27, which indicated a passable score in terms of usability. It was able to do most of its tasks, but it would require some improvements for it to reach an above average rating.