HUMANE-SE (Human-centric Mapping and Assessment of Needs and Expertise in Software Engineering) is a research-oriented framework designed to analyze the coexistence of technical and human-centric skills in Software Engineering job markets.
The project investigates how uniquely human skills — referred to as humics — remain essential in increasingly automated and AI-driven development environments.
This repository contains the methodology, documentation, and implementation concepts developed as part of an undergraduate thesis at the School of Informatics, Aristotle University of Thessaloniki.
The HUMANE-SE framework combines labor market data collection, ESCO-based skill extraction, semantic similarity mapping, network analysis, and AI-assisted recommendation generation to identify hybrid human-centric Software Engineering skill profiles.
The global co-occurrence network illustrates relationships between human-centric and technical skills identified in Software Engineering job postings.
Communication, adaptability, problem-solving, collaboration, and professional development emerged as some of the most frequently requested humics skills.
The HUMANE-SE platform includes an interactive humics exploration environment capable of generating job role and learning pathway recommendations using LLM-assisted workflows.
The main objectives of HUMANE-SE are:
- Identify human-centric skills in Software Engineering job postings
- Analyze the coexistence of humics with technical competencies
- Detect hybrid skill clusters through network analysis
- Generate recommendations for future job roles and learning pathways
- Explore how AI transforms Software Engineering skill requirements
The HUMANE-SE framework consists of four major stages:
- Data Collection
- Skill Extraction
- Data Analysis
- Humics Suggestion
The framework combines:
- ESCO taxonomy
- NLP-based skill extraction
- Semantic similarity mapping
- Co-occurrence network analysis
- Human-in-the-loop validation
Humics are uniquely human skills that remain resistant to automation, including:
- Communication
- Adaptability
- Empathy
- Creativity
- Critical thinking
- Ethical reasoning
- Collaboration
Humics Hubs represent clusters of co-occurring human-centric and technical skills identified through network analysis.
- Python
- ESCO / ESCOX
- NetworkX
- NLP / Transformer-based extraction
- Semantic similarity analysis
- Plotly / Graph visualization
- LLM-assisted recommendations
- Author: Aristotelis Kalantzis
- Institution: Aristotle University of Thessaloniki
- School: School of Informatics
- Software Engineering
- AI & Labor Market Analysis
- Human-centric Computing
- Skill Intelligence
Future extensions of the framework may include:
- Real-time labor market monitoring
- Multi-taxonomy skill mapping
- Explainable AI integration
- Expanded occupation similarity networks
- Cross-domain humics analysis
This project is licensed under the MIT License. See the LICENSE file for more information.