The task commonly referred to as spoken language understanding (SLU), or less misleadingly slot and intent detection (SID), consists in annotating prompts given to a digital assistant like Siri or Alexa. For example, the prompt "Set an alarm for 9am tomorrow" has the intent "add_alarm" and the date/time slot "9am tomorrow". The xSID dataset contains test data for a dozen languages and non-standard varieties like South Tyrolean German and Neapolitan. We are currently recruiting annotators to translate this dataset to a range of Norwegian dialects.
With your thesis, you could (a) contribute to these annotation efforts and (b) evaluate the cross-lingual transfer capabilities of various SID systems.
https://aclanthology.org/2021.naacl-main.197/
https://aclanthology.org/2023.vardial-1.25/
https://bitbucket.org/robvanderg/xsid/src/master/