Assessment of Learning technologies and Frameworks for Intelligent and Ethical AI

Informazioni relative al progetto

ALFIE

ID dell’accordo di sovvenzione: 101177912

DOI

10.3030/101177912

Data della firma CE 23 Settembre 2024

Data di avvio 1 Ottobre 2024

Data di completamento 30 Settembre 2027

Finanziato da

• Culture, creativity and inclusive society

Costo totale

€ 2 999 882,50

Contributo UE

€ 2 999 882,50

Coordinato da

CATALINK LIMITED
Cyprus

CORDIS fornisce collegamenti ai risultati finali pubblici e alle pubblicazioni dei progetti ORIZZONTE.

I link ai risultati e alle pubblicazioni dei progetti del 7° PQ, così come i link ad alcuni tipi di risultati specifici come dataset e software, sono recuperati dinamicamente da .OpenAIRE .

Risultati finali

Documents, reports (8)

Initial market analysis, exploitation plan and business models

The initial version of this deliverable will include an updated version of the market analysis and the exploitation plan, as well as specific business models from project assets.

Initial AI tools for multilingual interaction and automatic generation of AI models

This deliverable will present the initial AI tools that will be produced from automatic generation of AI models and multilingual interaction with the end-users.

Initial use case definition and AutoML validation

The initial version of this deliverable will describe the 3 pilot use cases, elaborating on the challenges, needs and demonstration approaches and will outline the evaluation methodology to be pursued.

Project website, market analysis, dissemination and exploitation plan

This deliverable will showcase the project website, present the initial dissemination plan of the project and the project’s communication material. It will also contain a detailed market analysis of AI and related technologies domain, the industrial requirements needed to consider, and the initial exploitation plan for each component of ALFIE system

Mid-term use case definition and AutoML validation

The mid-term version of this report will present the evaluation metrics of the trial use cases and new requirements for consideration by the AutoML tools.

Initial modern AI deployments and ethical gaps

D2.1 will provide a comprehensive report presenting a balanced perspective on the modern AI technologies and gaps

Initial best practices blueprint for ethical AI and AutoML Integration

This deliverable aims to consolidate a comprehensive best practices blueprint that will serves as a guide for designing an ethical AI framework and seamlessly integrating EU policy recommendations into the AutoML platform

Initial AutoML data warehouse and semantic knowledge graph

This deliverable will report the initial data collected from the end-users and pilot sites, as well as the retrospective data collected from other sources. Details of the initial semantic knowledge graph and semantic reasoner will also be included.

Data Management Plan (1)

Initial data management plan

This deliverable will describe the data generated in the project and will categorise them with respect to (i) their confidentiality, ethical treatment, possibility to be shared in the open data initiative, and (ii) their format and standardisation. Finally, it will specify guidelines for the management of all generated data in accordance with their categorization, as well as the EU ethical frameworks.

Pubblicazioni

Peer reviewed articles (5)

Ethical AI prompt recommendations in large language models using collaborative filtering

Autori: Jordan Nelson, Almas Baimagambetov, Konstantinos Avgerinakis, Nikolaos Polatidis
Pubblicato in: International Journal of Parallel, Emergent and Distributed Systems, 2025, ISSN 1744-5760
Editore: Informa UK Limited
DOI: 10.1080/17445760.2025.2573086

Cluster-specific localized drift detection for efficient batch model adaptation under controlled distribution shift

Autori: Ignacio Cabrera Martin, Marcello Trovati, Almas Baimagambetov, Nikolaos Polatidis
Pubblicato in: Expert Systems with Applications, Numero 331, 2026, ISSN 0957-4174
Editore: Elsevier BV
DOI: 10.1016/J.ESWA.2026.133370

Leveraging Ethical Narratives to Enhance <scp>LLM</scp>‐<scp>AutoML</scp> Generated Machine Learning Models

Autori: Jordan Nelson, Michalis Pavlidis, Andrew Fish, Nikolaos Polatidis, Yannis Manolopoulos
Pubblicato in: Expert Systems, Numero 42, 2025, ISSN 0266-4720
Editore: Wiley
DOI: 10.1111/EXSY.70072

AI-driven sign language interpreting avatars: navigating challenges to address communication gaps

Autori: Estella Oncins, Iris Serrat-Roozen, Pilar Orero
Pubblicato in: Universal Access in the Information Society, Numero 25, 2026, ISSN 1615-5289
Editore: Springer Science and Business Media LLC
DOI: 10.1007/S10209-026-01388-7

When AI fails to listen: convergent failure patterns across speech impairments and low-resource speech in ASR

Autori: Reyhaneh Sohrabi
Pubblicato in: AI &amp; SOCIETY, 2026, ISSN 0951-5666
Editore: Springer Science and Business Media LLC
DOI: 10.1007/S00146-026-03073-Y

Conference proceedings (5)

Deterministic and Trustworthy LLM-Driven Semantic Knowledge Graph Construction

Autori: Chukwudi Uwasomba, Nonso Nnamoko, Ioannis Korkontzelos
Pubblicato in: 2026 International Conference on Circuit, Systems and Communication (ICCSC), 2026
Editore: IEEE
DOI: 10.1109/ICCSC71566.2026.11650062

Utilizing Large Language Models for Machine Learning Explainability

DOI: 10.48550/ARXIV.2510.06912

Artificial Intelligence Driven Sign Language Interpreting Avatars

Autori: Estel·la Oncins, Pilar Orero & Iris Serrat
Pubblicato in: Kaleidoscope, Numero 2026, 2026, ISBN 978-92-61-42481-7
Editore: ITU
DOI: 10.5281/ZENODO.21496345

Behaviourally Informed Adversarial Prompting for Evaluating Emergent Risk Behaviours in Foundation Models

Autori: Chukwudi Uwasomba, Nonso Nnamoko, Ioannis Korkontzelos
Pubblicato in: 2026 International Conference on Circuit, Systems and Communication (ICCSC), 2026
Editore: IEEE
DOI: 10.1109/ICCSC71566.2026.11650133

Leveraging Generative AI, Knowledge Graphs and Personalised Learning Paths to Improve the Learning Experience of Neurodivergent Students in E-Learning Courses

Autori: Giacomo Nalli, Yoshi Malaise, Stelios Kapetanakis, Beat Signer
Pubblicato in: Lecture Notes on Data Engineering and Communications Technologies, Advanced Information Networking and Applications, 2026, ISBN 978-3-032-23260-1
Editore: Springer Nature Switzerland
DOI: 10.1007/978-3-032-23260-1_20

Monographic books (1)

ETHICOMP 2025 22nd international conference on the ethical and social impacts of ICT Lisbon

Autori: Isabel Alvarez, Adrian Didiu
Pubblicato in: 2025
Editore: Autónoma Edições
DOI: 10.26619/978-989-9002-62-3

Book chapters (1)

Artificial intelligence in Europe: challenges and opportunities towards communication for all.

Autori: Serrat-Roozen, I, P. Orero and E.Oncins
Pubblicato in: Tecnologías emergentes y sociedad conectada: automatización, datos y mediaciones contemporáneas., 2026, ISBN 979-13-7038-518-7
Editore: Tirant lo Blanc