Adapting existing and developing new open-source tools for specific user tasks
- text classification: e.g. sentiment classification
- document authentication: author profiling, authorship attribution, authorship verification
- automatic student assessment (grading of text)
- linguistic analysis: e.g. named entity recognition (NER), grammatical parsing
- topic modeling and trend analysis
- ...
Designing pipelines of existing and new tools for specific tasks
Applying explainable AI tools to specific tasks
- feature analysis: which information does the model exploit?
- model inspection: does the model behave in a robust way?
- model criticism: what exactly is the model good or bad at?