About CS Expo

CS Expo 2025.2.0: Voyager’s Odyssey is a celebration of innovation and creativity within FEU Institute of Technology. It serves as the premier platform for Computer Science students from Software Engineering and Data Science to present their thesis projects, demonstrating technical expertise and delivering solutions that address real-world challenges.

Bringing together students, faculty, industry partners, and guests, this immersive event showcases the power of young minds in a rapidly evolving technological landscape. Digital Reverie invites visionaries and innovators alike to imagine, create, and build solutions that transform the world.

THE VOYAGERS

Nineteen teams proudly presented projects that reflected their creativity and commitment to solving practical problems.

BandWitt

Farmulate: Optimizing Crop Rotation and Companion Planting Through XGBoost and YOLOv8n Algorithms Based on Soil Analysis

This study addresses the lack of simple and accessible tools for soil analysis, crop rotation, and companion planting, especially for small-scale and beginner farmers. Titled “Farmulate: Optimizing Crop Rotation and Companion Planting Through Machine Learning and Image Processing Algorithms Based on Soil Analysis,” it aims to help users choose suitable crops and planting strategies using machine learning and image processing. Designed for farmers, agricultural practitioners, and learners, the system analyzes soil images and sensor-based nutrient data to identify soil type and recommend appropriate crops and companion plants. It is expected to improve soil health, increase crop yield, and reduce the need for expensive laboratory soil testing, contributing to more sustainable and data-driven farming practices.

3J+R

A Machine Learning Approach to Parking Systems with Bidirectional Search and A* Algorithm for Optimized Pathfinding

This study addresses the problem of inefficient parking search and route navigation in urban environments with limited real-time parking data. Titled “A Machine Learning Approach to Parking Systems with Bidirectional Search and A Algorithms for Optimized Pathfinding,”* it aims to recommend suitable parking areas and generate efficient routes using machine learning–based suitability ranking and graph-based pathfinding algorithms. Designed for urban drivers and parking system administrators, the system evaluates parking options based on distance, cost, safety, accessibility, and operating hours, while computing optimal routes using Bidirectional A*. It is expected to reduce parking search time, traffic congestion, and user decision effort, contributing to smarter, more sustainable urban mobility solutions.

PhoenixCycle Coders

Smart Recommendation System for Electric Consumption on Appliance-Level Data With an Arduino-Based Device for Data Gathering Using LSTM Forecasting Algorithm

The study addresses the need for appliance‑level energy monitoring to help users reduce unnecessary electricity consumption through third-party devices. The research “Smart Recommendation System for Electric Consumption on Appliance-Level Data With an Arduino-Based Device for Data Gathering Using LSTM Forecasting Algorithm” aims to forecast electric usage through LSTM algorithm and use it together with historical data through an Arduino-based device to provide energy‑saving suggestions with a recommendation system. Intended for households that uses electricity especially individuals seeking to monitor and optimize their energy use. Through data‑driven insights, the project contributes to sustainable energy management by offering a low‑cost, intelligent solution for appliance‑level monitoring and optimization.

AlgoOmega

REFocus:Screen-Locking Android Application Using Decision Tree for Managing Screen-Time Overuse

This study addresses of overuse of screentime. Titled “ReFocus”, it aims to Lessen Screentime USe using Decision Tree and Locking Mechanisms. Designed to limit screentime overuse for android users, the system Records screentime and decides if the user overuse it. It is expected to Lock app specified by the user, contributing to Lessen Screentime and improve general focus and productivity.

Kobo

Sentiment Analysis of Structured and Unstructured Employee Feedback in a Multi-Tenant Environment Using Machine Learning Algorithms.

This study addresses the difficulty in manually analyzing and quantifying large volumes of qualitative employee feedback to extract actionable organizational insights. Titled "Kobo Sentiment Analyzer", it aims to automate the evaluation of employee satisfaction and identifying key themes in questionnaire responses using Natural Language Processing (NLP) techniques including VADER for sentiment analysis and BERTopic for topic modeling. Designed for Human Resources (HR) departments and organizational managers, the system processes text-based feedback to visualize sentiment trends, correlations, and feature importance. It is expected to provide real-time, objective insights into workforce morale, contributing to data-driven decision-making for improving organizational culture and employee retention.

Page 1 / 4