Overview
This page collects the work developed by GSI within the project AMOR.
Proyecto financiado por el Ministerio de Asuntos Económicos y Transformación Digital y por la Unión Europea – NextGeneration EU,Proyecto financiado por la Unión Europea (NextGenerationEU), dentro del Programa UNICO I+D Cloud.
Project funded by the Spanish Ministry for Economic Affairs and Digital Transformation and by the European Union - NextGenerationEU within the Programme UNICO I+D Cloud.
Chrome Plugin
The plugin provides an easy way to analyze the moral values of the web page the user is reading. This is being used to improve users' literacy and empower users against phenomena such as disinformation, clickbait, and fake news.
You can install it through this link.
Source Code
Moral Value Detection
- Moral value detection API:
- Available demo for the Moral value detection API
- Demo link: https://moral-values-api.gsi.upm.es
- API documentation: https://moral-values-api.gsi.upm.es/docs
Semantic Ethical Glass Box
- Semantic Ethical Glass Box Server:
- Documentation: https://segb.readthedocs.io.
- Running SEGB for AMOR Experiments:
- Secured by authentication: https://amor-segb.gsi.upm.es
Ontologies
One of the main tasks within the AMOR Project is annotating text with several types of information, such as moral values, emotions, and sentiments. The project uses and defines a combination of different linked data vocabularies for this.
The project has developed the following ontologies representing moral values and value annotations in textual sources. Available at link.
The ontologies are listed below.
1. AMOR Ontology
URL: http://www.gsi.upm.es/ontologies/amor/ns#
Description: This ontology provides the foundational concepts and relationships for Moral Annotation in the AMOR project.
2. AMOR Examples
URL: http://www.gsi.upm.es/ontologies/amor/examples#
Description: This ontology provides examples and use cases for annotating news in the AMOR project, demonstrating how the concepts and relationships can be applied.
3. Basic Human Values Taxonomy
URL: http://www.gsi.upm.es/ontologies/bhv/ns#
Description: This SKOS taxonomy mirrors the Basic Human Values Theory categories and provides a semantic representation using the standard SKOS representation.
4. AMOR Basic Human Values Ontology
URL: http://www.gsi.upm.es/ontologies/amor/models/bhv/ns#
Description: AMOR-BHV is an ontology that uses AMOR ontology and BHV SKOS Taxonomy to create the required relations for BHV to define moral annotations.
5. Moral Foundations Theory Taxonomy
URL: http://www.gsi.upm.es/ontologies/mft/ns#
Description: This SKOS taxonomy mirrors the categories in Moral Foundations Theory and provides a semantic representation using the standard SKOS representation.
6. AMOR Moral Foundations Ontology
URL: http://www.gsi.upm.es/ontologies/amor/models/mft/ns#
Description: AMOR-MFT is an ontology that uses AMOR ontology and MFT SKOS Taxonomy to create the required relations for using MFT to define moral annotations.
7. AMOR Experiments Ontology
URL: http://www.gsi.upm.es/ontologies/amor/experiments/ns#
Description: This ontology is designed to represent experiments and evaluations in the AMOR project. It includes classes and properties for defining experiments, experimentation subjects, parameters, and various strategies.
8. AMOR Experiments Examples
URL: http://www.gsi.upm.es/ontologies/amor/experiments/examples#
Description: This ontology provides examples and use cases for the AMOR Experiments Ontology, demonstrating how the concepts and relationships can be applied in experimental scenarios.
9. AMOR Datasets Examples Folder
Description: This folder contains some examples for NewsDataset used in the experiments.
10. SEGB Ontology
URL: http://www.gsi.upm.es/ontologies/segb/ns#
Description: This ontology provides the foundational concepts and relationships for the Semantic Ethical Glass Box (SEGB), inside the AMOR project, focusing on semantic representation of social and ethical guidelines for behavior.
11. SEGB Examples
URL: http://www.gsi.upm.es/ontologies/segb/examples#
Description: This ontology provides examples and use cases for the SEGB ontology, demonstrating how the concepts and relationships can be applied in various scenarios.
Publications
- Álvaro Carrera Barroso, José Luis Benítez Santana, Sergio Muñoz López & Carlos A. Iglesias (2025). Why was the AI agent biased? Towards a Semantic Ethical Glass Box for Social Robot and Avatar Auditing. In Proceedings of AIMMES 2025. Barcelona, Spain.
- Sergio Muñoz López & Carlos A. Iglesias. (2024). Exploiting Content Characteristics for Explainable Detection of Fake News. Big Data and Cognitive Computation, 8 (4), 1-18. Available at link.
- José Luis Benítez Santana. (2024). Design and Development of an Ethical and Moral Values Audit Toolkit for Detecting Bias and Fairness in Machine Learning-based Text Classifiers. Trabajo Fin de Titulación (TFM). Universidad Politécnica de Madrid, ETSI Telecomunicación. Available at link.
- Saulo José Nuez Ortega. (2024). Design and Development of a Semantic Ethical Black Box for a Social Robot. Audit of the Bias and Fairness of Training Datasets. Trabajo Fin de Titulación (TFM). Universidad Politécnica de Madrid, ETSI Telecomunicación. Available at link.
- Pablo Fernández Fraile. (2024). Design and Development of a Browser Plugin for Analysing Moral Emotions based on Intelligent Text Analytics. Trabajo Fin de Titulación (TFG). Universidad Politécnica de Madrid, ETSI Telecomunicación. Available at link.
- Anny Álvarez Nogales. (2024). Development of a Moral Foundations Estimation System based on Natural Language Processing Techniques and Transformer Models. Trabajo Fin de Titulación (TFG). Universidad Politécnica de Madrid. Available at link.
- Rocío Jiménez Villén. (2024). Implementation and Evaluation of Knowledge Graph-Based Models for News Recommendation. Trabajo Fin de Titulación. Available at link.
- Daniel Russo, Oscar Araque & Marco Guerini (2024). To Click It or Not to Click It: An Italian Dataset for Neutralising Clickbait Headlines. In Proceedings of the Tenth Italian Conference on Computational Linguistics (CLiC-it 2024). Available at link.
- Beatriz Hernandez-Fonta Codesido. (2024). Development of a Fake News Detection System using Machine Learning and Natural Language Processing Techniques. Trabajo Fin de Titulación (TFG). Universidad P
Press
AMOR has been disseminated to the public in the following media:
- 15/04/24 Descubren cómo detectar de manera automática contenidos orientados a radicalizar, UPM NewPress, available here.
