Radical Innovations


The main objective of the "Radical Innovations" project is to generate a better understanding of how “radical innovations” originate and contribute to the performance of companies. Automated content recognition in text (in patents, scientific publications and business news) yields an important set of indicators.


The project is carried out in collaboration with the Faculty of Business and Economics of K.U.Leuven ( Prof. Reinhilde Veugelers). LIIR is involved in the text mining aspects of the project.


We have built software for document clustering that relies on probabilistic topic models. We have researched feature extraction and hierarchical classification of patents, which is a challenging task to perform automatically given the large amounts of classification codes (in the thousands) and the limited number of training examples.

Period From 2011-10-01 to 2016-09-30.
Financed by K.U.Leuven: GOA/12/003
Supervised by Marie-Francine Moens
Staff Wim De Smet
Juan Carlos Gomez
Contact Juan Carlos Gomez


  1. GOMEZ, Juan Carlos & MOENS, Marie-Francine Multilayered Class Discrimination in Large-Scale Taxonomies. In Proceedings of the 16th International Conference on Knowledge-Based and Intelligent Information & Engineering Systems (Frontiers in Artificial Intelligence Applications 243) (pp. 615-625). Amsterdam: IOS Press. 2012
  2. DE SMET, Wim & MOENS, Marie-Francine Representations for Multi-Document Event Clustering. Data Mining and Knowledge Discovery, 26(3): 533-558. 2013
  3. GOMEZ, Juan Carlos & MOENS, Marie-Francine Document Categorization Based on Minimum Loss of Reconstruction Information. In Proceedings of the 11th Mexican International Conference on Artificial Intelligence (MICAI 2012) (Lecture Notes in Computer Science). Berlin: Springer. 2012
  4. GOMEZ, Juan-Carlos & MOENS, Marie-Francine Minimizer of the Reconstruction Error for Multi-Class Document Categorization. Journal of Expert Systems with Applications, 41 (3), 779-930. 2014
  5. GOMEZ, Juan Carlos & MOENS, Marie-Francine A Survey of Automated Hierarchical Classification of Patents. In G. Paltoglou, F. Loizides H. and. Preben (Eds.), Professional Search in the Modern World (Lecture Notes in Computer Science). Springer. 2014
  6. GOMEZ, Juan Carlos, HOSKENS, Stijn & MOENS, Marie-Francine Evolutionary Learning of Meta-Rules for Text Classification. In Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2017). ACM. 2017

Back to all projects