Deep learning is now a buzz word in big data analysis given its recent successes in addressing data processing challenges, specially using images and videos. The high-performance of deep learning solutions in now also being explored by text analysis systems in multiple scenarios, being the Google translation service one of the most popular case of success.

This project intends to study deep learning together with distant supervision techniques to improve Named Entity Disambiguation (NED) solutions, more specifically by automatically identifying deep relations that may enrich current coherence graph model approaches.

Additionally, the project intends to explore the full structure of KBs together with the deep relations by using semantic measures for propagating similarity in the coherence graph models. These novel approaches will be assessed in different gold standard corpus using different KBs, including popular and highly complex biomedical ontologies and DBPedia.