Stockholm university

Amanda KannPhD Student

About me

PhD student in Computational Linguistics at the Department of Linguistics since 2023.

My PhD project floats in the boundary between linguistic typology and computational linguistics. I am investigating how computational methods can help linguists extract typological features from large parallel texts [1] and how these features can be applied in multilingual NLP (natural language processing). 
I have several other research interests – here are a few:

  • how features of informal language use, such as hyperbolic numerals like "a zillion", can be investigated cross-lingually with the help of large parallel texts [2];
  • how multilingualism and plurilingualism are addressed (or ignored) in human-computer interaction [3];
  • how language technology can be used in higher-education pedagogy research [4]

I teach courses in computational linguistics, programming and human-computer interaction. You can find more information about my current teaching in the Teaching section below.

I am also active in student influence at SU. Currently, I am vice chair of the Humanities Faculty Council, and PhD representative in the faculty-level working group for education issues (HUGO) and the board of the Centre for University Teaching (CeUL). Please send me an email if you'd like to know more about what these groups do – information in English can be difficult to come by!

I was formerly Project Administrator at Språkstudion – for questions regarding the Language Cafes and other projects related to Språkstudion, please contact info@sprakstudion.su.se.

Teaching

I teach courses at undergraduate and graduate level at the Department of Linguistics, primarily in computational linguistics.

During 2024, I am teaching in the following courses:

I also have an interest in course development and scholarship of teaching and learning, with a particular focus on assessment and examination.

Publications

A selection from Stockholm University publication database

  • Voice Assistants Have a Plurilingualism Problem

    2022. Amanda Kann. CUI '22

    Conference

    Intelligent personal assistants (IPAs) using speech interfaces have historically been limited to monolingual use in pre-selected languages. Although recent developments in some IPAs have allowed for increased multilingual flexibility, the plurilingual competence-ability to utilise more than one language in the frame of a single interaction-of state-of-The-Art IPAs still falls short. This is demonstrated in a pilot study, where two widely used IPAs are shown to consistently fail in plurilingual interactions across 3 core tasks. This lack of plurilingual competence makes certain IPA functions virtually unusable in various contexts for users who are not native speakers of the official language(s) where they are located, and also speaks to wider problems in the treatment of multilingual use of IPAs by developers. Addressing these issues will not only make IPAs with speech interfaces considerably more functional for a large demographic of current and potential IPA users, but also enable new applications for IPAs in contexts such as self-regulated language learning. 

    Read more about Voice Assistants Have a Plurilingualism Problem

Show all publications by Amanda Kann at Stockholm University