KONSTRAKT-An AI-Powered Job and Project Networking Platform for the Philippine Construction Industry Using TF-IDF and Cosine Similarity Matching
Year: 2026
Type: conference-paper
Abstract: The Philippine construction industry, despite being a key economic driver, faces persistent productivity challenges stemming from a severe skills mismatch and dependence on fragmented, informal recruitment networks. Approximately 80% of the sector's workforce operates informally, with an estimated 65–70% of local hiring conducted through word-of-mouth and the traditional "standby" system, where workers physically gather at street corners awaiting daily wage opportunities. These inefficiencies result in underemployment among skilled workers and recurring project delays for contractors, largely due to the absence of a centralized, credentialverified hiring channel. This study developed KONSTRAKT, a mobile and web-based job networking platform designed to professionalize construction recruitment in Metro Manila. Built on React Native for cross-platform accessibility and Supabase for a secure, scalable backend, the platform's core innovation is an AIpowered matching engine developed using Python and scikit-learn. The engine applies Term Frequency-Inverse Document Frequency (TF-IDF) and Cosine Similarity algorithms to structured user profiles, prioritizing verifiable TESDA certifications, technical skills, and geolocation data to generate accurate worker-to-project compatibility scores. By transitioning the industry from informal recruitment practices to a data-driven digital marketplace with verifiable credentials, KONSTRAKT reduces search friction, increases workforce visibility, and bridges the trust gap between workers and contractors. The platform advances inclusive economic growth aligned with the United Nations Sustainable Development Goals 8 (Decent Work and Economic Growth) and 9 (Industry, Innovation, and Infrastructure).
Institution: Lyceum of the Philippines University
Language: English
Topic: AI and HR Technologies
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7167478000