Some of the examples make use of the DaNLP pip package and show how to use the package. Other examples focus on different use cases e.g. applying cross-transfer learning on Danish.
Jupyter notebooks :
- Getting started tutorial, including examples of how to load and use DaNLP models and datasets (similar to the code snippets from our documentation) :
getting_started.ipynb - Tutorial of applying zero shot transfer learning to train a sentiment
classifier that can be applied on Danish text:
example_zero_shot_sentiment.ipynb - Tutorial about how to learn a text classification model,
here exemplified by a model that can predict which party a speech
(from the Danish parliament) belongs to :
example_text_classification.ipynb - Tutorial about how to perform data augmentation on Danish text,
here exemplified by augmenting Tweets and testing if the increased data
increases the performance of a sentiment model:
example_data_augmentation.ipynb - Tutorial about how to build a simple knowledge graph using NER, POS-tagging,
dependency parsing and coreference resolution :
example_knowledge_graph.ipynb