Projects

Machine Learning

Customer Purchase Propensity Analysis with Explainable AI

Uncovering the key drivers behind customer drop-offs in e-commerce is crucial to unlocking business growth. Using Explainable AI, our goal was to identify these drivers and pinpoint actionable strategies that will retain and convert customers.

Recommender System

In today's marketing landscape, the power of influencer marketing is undeniable. Yet, the meticulous process of choosing the right partners poses challenges both in terms of time and expense. To address this, we harnessed the potential of content-based recommender system and information retrieval techniques.

Natural Language Processing & Information Retrieval

Enhancing Service Support Through Customer Concern Topic Modelling

In a highly connected nation, a temporary Internet shutdown leads to reduced productivity, communication challenges, and a surge in customer complaints. With topic modeling techniques, we identify and categorize complaint characteristics that enable internet providers to swiftly allocate concerns to the right department, expediting resolutions and enhancing customer experience.

Deep Learning

Proactive Manufacturing Quality Control with Anomaly Detection & Explainable AI

In manufacturing, maintaining top-tier quality standards is essential. However, guaranteeing this quality is challenging due to the rarity, scale, sensitivity, and complexity of defects. By combining advanced data mining, deep learning techniques, and explainable AI, we address these challenges to help quality management for manufacturing companies.

Big Data & Cloud Computing

From Publication to Citation: Understanding the Factors Driving Research Recognition

With about 90% of journal articles remaining uncited, and around 50% unnoticed by authors and editors, maximizing the impact of scholarly work is important. Using big data analysis and machine learning, we identify factors that influence article citations and readership to guide researchers and institutions on which articles to focus based on their potential for impact.

Machine Learning Operations

Money Guard: Streamlining Financial Management with Computer Vision

In the Philippines, 94% of Filipinos still prefer cash transactions despite the growth of fintech. However, manual bill counting poses challenges like human error, fraud, inefficiency, and time consumption. To tackle these issues, we developed a computer vision application to streamline financial tasks, reducing human error and enhancing efficiency.

Publications

Simulating network formation through in-person interaction in a limited budget environment

Proceedings of the Samahang Pisika ng Pilipinas, 37 (2019)
Authors: Maria Loraine R. Menorca and May T. Lim

Proximity analysis between researchers and networking venues

Proceedings of the Samahang Pisika ng Pilipinas, 36 (2018)
Authors: Maria Loraine R. Menorca and May T. Lim