2025 · Preprint (arXiv)
PatenTEB: A Comprehensive Benchmark and Model Family for Patent Text Embedding
Iliass Ayaou, Denis Cavallucci
arXiv preprint
People / Founder and coordinator
Prof. Denis Cavallucci
Founder and coordinator of the Chair: political and scientific leadership
Research domains
- TRIZ & inventive design
- Computer-aided innovation / ontologies
- Problem formulation (problem graphs, contradictions)
- AI & NLP for invention (patents, LLMs)
- Creativity & innovation education
Denis Cavallucci works on the theory and tooling of inventive design, building on TRIZ and its extension OTSM-TRIZ. His work has moved from formalising inventive concepts (laws of system evolution, contradictions, problem graphs, ontologies for computer-aided innovation) to using artificial intelligence — patent mining, language models, ARIZ automation — to assist invention in industrial R&D.
Related projects
- Définition des indicateurs de l’efficience inventive pour caractériser les activités inventives en R&D : application au domaine de l’automobile
- Contribution à l’extraction de contradictions techniques depuis un corpus de brevets propre à un domaine par l’usage des techniques du traitement automatique des langues
- Text mining for automating TRIZ-based inventive design process using patent documents
- Inventive solutions retrieval from patent documents via natural language processing
- Amélioration du processus de la conception inventive par l’utilisation de méthodes agiles et d’algorithmes d’apprentissage automatique
- Experience capitalization to support inventive design studies
- Contribution à la formalisation des liens invention-optimisation en conception inventive
- Contribution à la méthode de conception inventive par l’extraction automatique de connaissances des textes de brevets d’invention
- (Meta)Knowledge modeling for inventive design
- Towards an innovative methodology for knowledge fusion from patent datasets: application to inventive design
- Text mining and topic modelling for innovation analysis
- Interdisciplinary knowledge for inventive problem solving
- Patent intelligence and LLM orchestration of ARIZ
Publications
2025 · Preprint (arXiv)
DAPFAM: A Domain-Aware Family-level Dataset to benchmark cross domain patent retrieval
Iliass Ayaou, Denis Cavallucci, Hicham Chibane
arXiv preprint
2025 · Conference paper
CPU-Efficient Verification of Science Problem–Solution Pairs: Design Rationale and Baselines
Nicolas Douard, Ahmed Samet, George Giakos, Denis Cavallucci
International TRIZ and Artificial Intelligence Conference
2025 · Conference paper
Prétraitement syntaxique pour enrichir le Bag of Words en Topic Modeling
Connor Maclean, Denis Cavallucci
20e Conférence en Recherche d’Information et Applications (CORIA) 32ème Conférence sur le Traitement Automatique des Langues Naturelles (TALN) 27ème Rencontre des Étudiants Chercheurs en Informatique pour le Traitement Automatique des Langues (RECITAL) Les 18e Rencontres Jeunes Chercheurs en RI (RJCRI)
2025 · Journal article
Quantifying Interdisciplinarity in Scientific Articles Using Deep Learning Toward a TRIZ-Based Framework for Cross-Disciplinary Innovation
Nicolas Douard, Ahmed Samet, George Giakos, Denis Cavallucci
Machine Learning and Knowledge Extraction
2025 · Conference paper
FusionTRIZ Framework for Cross-Disciplinary Innovation: A Neuromorphic Imaging Application
Nicolas Douard, Ahmed Samet, Denis Cavallucci, George Giakos
2025 IEEE International Conference on Imaging Systems and Techniques (IST)
2025 · Journal article
Large-scale transformer-based topic graphs identify thematic links between engineering and biology
Nicolas Douard, Denis Cavallucci, Ahmed Samet, George Giakos
Scientific Reports
2025 · Book
World Conference of AI-Powered Innovation and Inventive Design
Denis Cavallucci, Stelian Brad, Pavel Livotov
Related
Tools7
- MatrixGenGenerate domain-specific contradiction matrices from patents (research prototype).
- SummaTRIZResearch demonstrator from doctoral work (status to be confirmed).
- TRIZBenchFrom a problem statement to the few contradictions that really matter.
- HiveScanMap innovative companies of a field, with an innovation index you can decompose.
- FusionMindLook for the solution to an engineering problem in other scientific disciplines.
- PatentSparkFind, in patents, the solutions and contradictions related to your problem.
- ARIZ-CopilotARIZ executed step by step by language models, anchored in real patents.