The paper “Modeling Topic Dependencies in Hierarchical Text Categorization” discusses about automatically learning from the structure of a text categorization hierarchy to better classify documents.This is a step towards knowledge evolution since it shows how to learn the structure and thus possibly its evolution. Co-written by A. Moschitti, J. Qi and R. Johansson, it will be published in the Conference for the Association for Computational Linguistics, (ACL). Jeju, Korea, 2012.
The LK project is funded by the European Commission under Project No. 231126