University showcase15min
From Skills to Career Paths: Explainable Recommendations with Knowledge Graphs
This thesis develops an explainable career recommendation system using ESCO, knowledge graphs, semantic skill matching, graph algorithms, and labor-market data. It matches current skills to occupations, identifies skill gaps, suggests career paths and learning routes, and explains recommendations by showing the evidence behind each match.
talk.summaryAiDisclaimer
Ourania PapadopoulouAristotle University of Thessaloniki
talkDetail.whenAndWhere
Friday, November 20, 17:05-17:20
Room 3 - Alexandros
Choosing a career path is often difficult because recommendations may show a result without explaining why it fits a person’s current skills. This undergraduate thesis explores the development of an explainable career recommendation system that combines ESCO occupational and skill data, Knowledge Graphs, semantic skill matching, graph algorithms, and supporting labor-market evidence from Skillab.
The system compares a user’s current skills with occupation requirements, identifies skill gaps, suggests possible career directions, and provides personalized skill recommendations and learning paths. Rather than returning only a ranking or score, it also exposes the evidence behind each recommendation, helping users understand how their existing skills contribute to a career match and which additional skills could improve their fit.
The presentation gives a short overview of the system’s architecture, recommendation approach, and explainability mechanisms.
The system compares a user’s current skills with occupation requirements, identifies skill gaps, suggests possible career directions, and provides personalized skill recommendations and learning paths. Rather than returning only a ranking or score, it also exposes the evidence behind each recommendation, helping users understand how their existing skills contribute to a career match and which additional skills could improve their fit.
The presentation gives a short overview of the system’s architecture, recommendation approach, and explainability mechanisms.
Ourania Papadopoulou
I'm a Computer Science student and backend developer. I build secure APIs and architecture mostly in .NET/C#, but when it comes to data and graphs, I switch to Python. Currently, I am researching and developing advanced career recommendation systems leveraging Neo4j graph databases, semantic embeddings, and ESCO taxonomies. I am passionate about building secure architectures and exploring the intersection of graph data with practical software solutions. Always up for a chat about backend architecture, API security, or graph data.