KDDE - Knowledge Discovery and Data Engineering

01/05/2026

📢 𝐀 𝐃𝐀𝐓𝐀‑𝐂𝐄𝐍𝐓𝐑𝐈𝐂 𝐕𝐈𝐄𝐖 𝐎𝐍 𝐀𝐑𝐓𝐈𝐅𝐈𝐂𝐈𝐀𝐋 𝐈𝐍𝐓𝐄𝐋𝐋𝐈𝐆𝐄𝐍𝐂𝐄 📢

The KDDE - Knowledge Discovery and Data Engineering is pleased to share a recent research outcome produced within the 𝐅𝐀𝐈𝐑 𝐩𝐫𝐨𝐣𝐞𝐜𝐭, in the framework of 𝐓𝐏𝟕 – 𝐃𝐚𝐭𝐚‑𝐜𝐞𝐧𝐭𝐫𝐢𝐜 𝐀𝐈 𝐚𝐧𝐝 𝐈𝐧𝐟𝐫𝐚𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞𝐬.

An 𝐨𝐩𝐞𝐧‑𝐚𝐜𝐜𝐞𝐬𝐬 𝐬𝐜𝐢𝐞𝐧𝐭𝐢𝐟𝐢𝐜 𝐚𝐫𝐭𝐢𝐜𝐥𝐞 has just been published, evolving a previously released white paper into a full research contribution that serves as a 𝐃𝐚𝐭𝐚‑𝐂𝐞𝐧𝐭𝐫𝐢𝐜 𝐀𝐈 𝐌𝐚𝐧𝐢𝐟𝐞𝐬𝐭𝐨:

👉 https://www.mdpi.com/2079-9292/15/9/1913

The paper advocates a paradigm shift from 𝐦𝐨𝐝𝐞𝐥‑𝐜𝐞𝐧𝐭𝐫𝐢𝐜 𝐀𝐈 to 𝐝𝐚𝐭𝐚‑𝐜𝐞𝐧𝐭𝐫𝐢𝐜 𝐀𝐈. While traditional approaches focus on continuously changing models trained on mostly static datasets, the data‑centric perspective reverses this dynamic:
𝐦𝐨𝐝𝐞𝐥𝐬 𝐛𝐞𝐜𝐨𝐦𝐞 𝐜𝐨𝐦𝐩𝐚𝐫𝐚𝐭𝐢𝐯𝐞𝐥𝐲 𝐬𝐭𝐚𝐛𝐥𝐞, 𝐰𝐡𝐢𝐥𝐞 𝐝𝐚𝐭𝐚 𝐚𝐫𝐞 𝐜𝐨𝐧𝐭𝐢𝐧𝐮𝐨𝐮𝐬𝐥𝐲 𝐜𝐮𝐫𝐚𝐭𝐞𝐝, 𝐞𝐧𝐫𝐢𝐜𝐡𝐞𝐝, 𝐠𝐨𝐯𝐞𝐫𝐧𝐞𝐝, 𝐚𝐧𝐝 𝐢𝐦𝐩𝐫𝐨𝐯𝐞𝐝 throughout the AI lifecycle.

The work provides:
• a 𝐦𝐞𝐭𝐡𝐨𝐝𝐨𝐥𝐨𝐠𝐢𝐜𝐚𝐥 𝐚𝐧𝐝 𝐜𝐨𝐧𝐜𝐞𝐩𝐭𝐮𝐚𝐥 𝐟𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 for Data‑centric AI;
• a clear 𝐜𝐨𝐧𝐧𝐞𝐜𝐭𝐢𝐨𝐧 𝐰𝐢𝐭𝐡 𝐭𝐨𝐨𝐥𝐬, 𝐢𝐧𝐟𝐫𝐚𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞𝐬, 𝐚𝐧𝐝 𝐅𝐀𝐈𝐑 𝐝𝐚𝐭𝐚 𝐩𝐫𝐢𝐧𝐜𝐢𝐩𝐥𝐞𝐬;
• an up‑to‑date discussion of 𝐃𝐚𝐭𝐚‑𝐜𝐞𝐧𝐭𝐫𝐢𝐜 𝐀𝐈 𝐢𝐧 𝐭𝐡𝐞 𝐞𝐫𝐚 𝐨𝐟 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈, with emphasis on robustness, reliability, and responsible deployment.
This contribution highlights how 𝐝𝐚𝐭𝐚 𝐪𝐮𝐚𝐥𝐢𝐭𝐲 𝐚𝐧𝐝 𝐝𝐚𝐭𝐚 𝐩𝐫𝐨𝐜𝐞𝐬𝐬𝐞𝐬 𝐚𝐫𝐞 𝐤𝐞𝐲 𝐝𝐫𝐢𝐯𝐞𝐫𝐬 𝐨𝐟 𝐦𝐨𝐝𝐞𝐫𝐧 𝐀𝐈 𝐬𝐲𝐬𝐭𝐞𝐦𝐬, offering a reference framework for researchers, practitioners, and institutions.

𝐀𝐮𝐭𝐡𝐨𝐫𝐬
Donato Malerba
Antonella Poggi
Mario Alviano
Tommaso Boccali
Maria Teresa Camerlingo
Roberto Maria Delfino
Domenico Diacono
Domenico Elia
Vincenzo Pasquadibisceglie
Mara Sangiovanni
Vincenzo Spinoso
Gioacchino Vino

18/03/2026

🚀 𝗡𝗲𝘄 𝗢𝗽𝗲𝗻 𝗔𝗰𝗰𝗲𝘀𝘀 𝗣𝘂𝗯𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗯𝘆 𝗞𝗗𝗗𝗘 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗚𝗿𝗼𝘂𝗽!

Our new article — authored by 𝗖𝗼𝗿𝗿𝗮𝗱𝗼 𝗟𝗼𝗴𝗹𝗶𝘀𝗰𝗶, 𝗩𝗶𝘁𝗼 𝗡. 𝗟𝗼𝘀𝗮𝘃𝗶𝗼, 𝗦𝗮𝘃𝗲𝗿𝗶𝗼 𝗣𝗮𝘀𝗰𝗮𝘇𝗶𝗼 𝗮𝗻𝗱 𝗗𝗼𝗻𝗮𝘁𝗼 𝗠𝗮𝗹𝗲𝗿𝗯𝗮 — has just been published Open Access in 𝑸𝒖𝒂𝒏𝒕𝒖𝒎 𝑴𝒂𝒄𝒉𝒊𝒏𝒆 𝑰𝒏𝒕𝒆𝒍𝒍𝒊𝒈𝒆𝒏𝒄𝒆 (Springer).

🔗 𝗥𝗲𝗮𝗱 𝘁𝗵𝗲 𝗽𝗮𝗽𝗲𝗿 𝗵𝗲𝗿𝗲:
https://link.springer.com/article/10.1007/s42484-026-00374-9

This work introduces 𝗤𝗨𝗥𝗜𝗢𝗦𝗢, a framework that helps machine learning models become more efficient and reliable when dealing with real‑world, continuously evolving data. Instead of re‑optimizing quantum circuit parameters from scratch each time, QURIOSO predicts good parameters using classical or quantum‑enhanced LSTM models—leading to faster, more stable learning.

Why does this matter?
Because in many real applications, from streaming data to dynamic environments, models need to adapt quickly. Making quantum‑enhanced AI systems more robust and less dependent on costly re‑optimization is a key step toward practical quantum machine learning.

We’re proud to share this contribution with the scientific community.

In Variational Quantum Algorithms (VQAs), circuit parameters are typically re-optimized from scratch for each new dataset, an approach that becomes ineffic

06/03/2026

📢 New press release from the Università degli Studi di Bari Aldo Moro

🌳 An Artificial Intelligence model for satellite image analysis to monitor forest health.

The study proposes an AI-based approach to analyze satellite imagery and support forest monitoring, contributing to ecosystem protection and the sustainable management of natural resources.

🔗 https://www.uniba.it/it/ateneo/rettorato/ufficio-stampa/comunicati-stampa/anno-2026/modello-intelligenza-artificiale-analisi-immagini-satellitari-tato-salute-foreste

The research activity was supported by the SWIFTT Project and the PNRR project FAIR – Future AI Research.

Researchers involved in the study:
Vito Recchia, Giuseppina Andresini, Annalisa Appice, Dino Ienco, Giuseppe Fiameni, Donato Malerba.

5 marzo 2026

25/02/2026

📣 PhD Award Announcement – Luca De Rose

We are pleased to announce that today Luca De Rose, PhD candidate of the KDDE – Knowledge Discovery and Data Engineering Laboratory, successfully completed his PhD in Computer Science and Mathematics.

Luca brilliantly defended his dissertation entitled:

“Advancing Adversarial Learning with Explainable Artificial Intelligence in Cybersecurity”

Candidate: Luca De Rose
Supervisor: Prof. Annalisa Appice
Co-supervisor: Dr. Giuseppina Andresini

The Examination Committee commended the scientific quality and relevance of his research, which addresses a highly topical area at the intersection of adversarial learning, explainable AI, and cybersecurity.

We extend our warmest congratulations to Luca on behalf of the KDDE and the entire Department of Computer Science.

18/02/2026

Open Access Publication: Leveraging Triplet Autoencoders with Quantum Tensor Networks for Multiclass Image Classification

We are pleased to announce that the paper “Leveraging Triplet Autoencoders with Quantum Tensor Networks for Multiclass Image Classification”, authored by Vito Nicola Losavio, Maria Grazia Miccoli, and Donato Malerba from the KDDE, has been published in open access within the CEUR Workshop Proceedings (Vol. 4153).

The contribution was presented at the AIQ×QIA 2025 – 3rd International Workshop on AI for Quantum and Quantum for AI, and proposes a novel hybrid quantum–classical pipeline for multiclass image classification. The work integrates Triplet Autoencoders with quantum tensor network–inspired variational circuits, demonstrating improved performance on MNIST, especially in low‑data scenarios.

This research highlights the potential of combining structured latent representations with efficient quantum models in the NISQ era.

📄 Open Access Paper:
https://ceur-ws.org/Vol-4153/paper19.pdf

📚 Proceedings Volume:
https://ceur-ws.org/Vol-4153/

17/02/2026

We are pleased to announce the publication of the article:

Handling complex backgrounds and light perturbations for enhancing learning tasks from images of vegetables

by Stefano Polimena, Gianvito Pio, Giovanni Attolico, and Michelangelo Ceci

Published in the Journal of Intelligent Information Systems, Volume 64 (2026)

🔗 Open access: https://link.springer.com/article/10.1007/s10844-025-00984-y

Overview:
Assessing the quality of fruits and vegetables is a key step in the agro-food supply chain, but real-world images often suffer from complex backgrounds and varying lighting conditions. These factors significantly hinder the performance of automatic learning systems.
This work introduces a robust and effective preprocessing and feature extraction pipeline designed to make image-based learning tasks more reliable and less sensitive to visual perturbations.

Main Contributions:
- Background removal using a pre-trained U2-Net architecture
- Novel outlier‑detection procedure to eliminate residual background artifacts, particularly along product edges
- Extraction of complete color histograms
- Use of an autoencoder to learn high-level color representations at multiple levels of granularity
- Enhanced robustness to light and color variations in real-world scenarios

Results:
Experiments conducted on two real-world datasets demonstrate that the proposed solution:
✔️ Outperforms baseline and state‑of‑the‑art approaches
✔️ Improves stability under uncontrolled acquisition conditions
✔️ Provides a lightweight and generalizable framework for agro-food quality assessment

The quality assessment of fruits and vegetables is crucial in the agroalimentary supply chain, as it directly affects consumer satisfaction, market value and overall food security. Traditional approaches rely on visual inspections or destructive techniques, which are labor-intensive and time-consumi...

16/02/2026

🌟 New KDDE Research Publication in a Top‑Tier Q1 Journal!

We are thrilled to announce our latest Open Access article published in Information Systems (Vol. 139, July 2026) — a prestigious Q1 journal in Computer Science.

🔬 “Multimodal predictive process monitoring and its application to explainable clinical pathways”

by by Vincenzo Pasquadibisceglie, Ivan Donadello, Annalisa Appice, Oswald Lanz, Fabrizio Maria Maggi, Giuseppe Fiameni, and Donato Malerba

📘 What this research delivers
The paper presents MEDUSA, an innovative multimodal AI framework that integrates:
• 🩻 radiological images
• 📋 Electronic Health Records (EHR)
• ✍️ clinical text notes
Applied to a real COVID‑19 dataset, MEDUSA predicts mortality risk and provides transparent explanations using Integrated Gradients — helping clinicians understand how each modality contributes to the model’s decisions.

✨ Why this matters
This work advances the frontier of explainable multimodal AI in healthcare, demonstrating how heterogeneous clinical information can be fused to support decision‑making, enhance patient care, and strengthen trust in AI‑based predictions.

🔓 The article is fully Open Access — freely available to everyone.

📄 Read it here:
https://www.sciencedirect.com/science/article/pii/S0306437926000128

🔗 DOI: https://doi.org/10.1016/j.is.2026.102698

15/02/2026

Nuova pubblicazione – Contributo del Dr. Filippo Lorè su IA, trasparenza algoritmica e “riserva di umanità” nella PA

Il KDDE Research Group è lieto di segnalare il contributo del Dr. Filippo Lorè, pubblicato nel numero 1/2026 della rivista Diritto di Internet.

Il lavoro, intitolato “Accesso documentale e Intelligenza Artificiale: l’affermazione della ‘riserva di umanità’ nell’era digitale”, analizza una recente e rilevante decisione del Consiglio di Stato (Sez. VI, 5 giugno 2025), che affronta il tema dell’accesso agli atti amministrativi quando il procedimento decisionale è supportato da sistemi di Intelligenza Artificiale.

Il contributo approfondisce:
- il principio di trasparenza e il diritto di conoscere la logica decisionale degli algoritmi utilizzati dalla Pubblica Amministrazione;
- il ruolo imprescindibile della supervisione umana (human oversight) nei procedimenti digitali;
- i rischi di non discriminazione algoritmica e la necessità di rendere comprensibili e accessibili i criteri utilizzati dai sistemi di AI;
- l’impatto della sentenza sulla tutela dei diritti fondamentali e sulle nuove responsabilità della PA nell’era delle decisioni automatizzate.

La pubblicazione offre una riflessione ampia e rigorosa sul rapporto tra innovazione tecnologica e principi costituzionali, contribuendo al dibattito attuale sulla regolazione dell’IA nel settore pubblico, anche alla luce dell’AI Act e della recente legge italiana n. 132/2025.

📄 Riferimento bibliografico:
F. Lorè, “Accesso documentale e Intelligenza Artificiale: l’affermazione della ‘riserva di umanità’ nell’era digitale”, Diritto di Internet, n. 1/2026, pp. 171–183.

https://dirittodiinternet.it/wp-content/uploads/2026/01/abstract-1-26.pdf

05/02/2026

📢 New Publication from the KDDE Research Group

We’re excited to share that our latest OPEN ACCESS publication in the journal Quantum Machine Intelligence (Springer):

📄 “Leveraging cloud-native infrastructure for dynamic and flexible quantum-classical MLOps”
👩‍🔬 Authors: Angelo Impedovo, Vito Nicola Losavio & Corrado Loglisci
🔗 https://link.springer.com/article/10.1007/s42484-026-00360-1

This work presents a conceptual framework for integrating cloud-native infrastructures with hybrid quantum-classical MLOps, offering a flexible and scalable approach to managing machine learning and quantum workflows in contemporary computing environments.

📚 If your interests include Quantum Machine Intelligence, MLOps, or hybrid computing systems, give it a read!

Quantum Machine Learning (QML) approaches introduce unique computational requirements that continuously emerge, from data encoding to model training, from

04/02/2026

📢 New Scientific Publication from KDDE (Open Access)!

The article “ANAKIN: explainable Android malware detection with graph neural networks” has been published in Cybersecurity, a leading international journal in the field of cybersecurity by Springer Nature.

The paper presents ANAKIN, an innovative approach to Android malware detection that combines Graph Neural Networks (GNNs) with Explainable Artificial Intelligence techniques. The proposed method relies on graph-based representations of APIs extracted from decompiled APKs and integrates GNNExplainer to provide transparent and interpretable explanations of malware detection decisions.

📊 An extensive experimental evaluation conducted on more than 26,000 Android APKs demonstrates that ANAKIN outperforms traditional deep learning approaches, while also supporting root cause analysis of malicious behaviors.

👩‍🔬👨‍🔬 Authors: Giuseppina Andresini, Annalisa Appice, Vincenzo Belvedere, Giuseppe Fiameni, Donato Malerba

👉 https://link.springer.com/article/10.1186/s42400-026-00552-z

Android OS is today the most used Operating System for mobile devices. However, it is susceptible to several malware attacks that may seriously compromise the privacy and security of individuals and organizations. This paper proposes an approach based on a static analysis of decompiled Android PacKa...