In our work on user-tailored environmental information production, we are researching machine learning-based generation models, which require so-called "treebanks", i.e., corpora annotated with linguistic structures, for training.

Treebanks are widely used in natural language processing and computational linguistics applications. We offer the treebank resources created in PESCaDO as a service to the community.

For illustration of the types of information annotated in the Finnish treebank, see the following examples.



To obtain the treebanks for English and/or Finnish, please contact leo.wanner@upf.edu