@inproceedings{saynova-etal-2023-class,
    title = "Class Explanations: the Role of Domain-Specific Content and Stop Words",
    author = "Saynova, Denitsa  and
      Bruinsma, Bastiaan  and
      Johansson, Moa  and
      Johansson, Richard",
    editor = {Alum{\"a}e, Tanel  and
      Fishel, Mark},
    booktitle = "Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa)",
    month = may,
    year = "2023",
    address = "T{\'o}rshavn, Faroe Islands",
    publisher = "University of Tartu Library",
    url = "https://preview.aclanthology.org/ingest_wac_2008/2023.nodalida-1.12/",
    pages = "103--112",
    abstract = "We address two understudied areas related to explainability for neural text models. First, \textit{class explanations}. What features are descriptive across a class, rather than explaining single input instances? Second, the \textit{type of features} that are used for providing explanations. Does the explanation involve the statistical pattern of word usage or the presence of domain-specific content words? Here, we present a method to extract both class explanations and strategies to differentiate between two types of explanations {--} domain-specific signals or statistical variations in frequencies of common words. We demonstrate our method using a case study in which we analyse transcripts of political debates in the Swedish Riksdag."
}