Generating structured content is hard, often involving complex or unseen rules. This includes machine writing of documents, translations, summaries, or the general problem of putting items into order.

Deep Learning Generative Models for Content Structuring

InnovationsFonden, total 1.07M DKK, 2019-2022

The aim of this project is to create deep learning algorithms for generation of structured content, which take the vast amount of individual data points and combine them into relevant, engaging content.

• Summarizing scientific literature on the basis of deconstructed systematic reviews and meta-analyses * PROCAT: Product Catalogue Dataset for Implicit Clustering, Permutation Learning and Structure Prediction. 2021.
• Set-to-Sequence Methods in Machine Learning: a Review. 2021.

• Robotjournalistik skaber arbejde til it-folk - Prosa (Denmark), 2019.09.09