Hello world! ⚡ I am Myrthe Reuver, a PhD candidate at the Computational Linguistics and Text Mining Lab (CLTL) at the Vrije Universiteit Amsterdam (VU). My research is about news recommender systems and specifically: diversity in news recommender systems. I use NLP to explore and promote diversity, especially that of different viewpoints, in news recommendation contexts. I’m supervised by prof. dr. Antske Fokkens (also connected to CLTL) and dr. Suzan Verberne (connected to LIACS at Leiden University).

My specialization is specifically in Computational Argumentation or Argument Mining. I also have a keen interest in societal and cultural implications of NLP and AI, and interdisciplinary collaboration. I have extensive knowledge of current developments in NLP, and am extremely curious! My experience ranges from identifying research problems to managing annotation projects, and from exploratory data analysis to training AI models, including techniques such as few-shot learning and prompting. I am especially interested in careful evaluation: Are we measuring what we think we are measuring? 🤔 I also enjoy presenting research outcomes to different audiences - whether in talks, papers, or reports.

Before my PhD, I completed the Research Master’s programme “Linguistics and Communication Sciences” (cum laude + with Honours) at Radboud University in Nijmegen and the MSc programme “Cognitive Science and Artificial Intelligence” (with distinction) at Tilburg University in Tilburg. I also was a proud co-founder and editor-in-chief of Radboud’s student linguistics journal RU:ts, and enjoy teaching: I was TA on Information Science during my MA and have taught ‘Programming in Python’ as well as a research seminar on Argument Mining during my PhD. You can download my full CV here.

I hope to continue doing interesting research in the future, where my interests lie in societally and scientifically responsible NLP in complex domains! Want to know more about any of my current or past projects? Ask me through email, LinkedIn, or Twitter. I’m always interested in new ideas or opportunities, so please contact me if you have any! 😁

News

_{19 Sept 2023 Our work “𝘐𝘮𝘱𝘳𝘰𝘷𝘪𝘯𝘨 𝘢𝘯𝘥 𝘌𝘷𝘢𝘭𝘶𝘢𝘵𝘪𝘯𝘨 𝘵𝘩𝘦 𝘋𝘦𝘵𝘦𝘤𝘵𝘪𝘰𝘯 𝘰𝘧 𝘍𝘳𝘢𝘨𝘮𝘦𝘯𝘵𝘢𝘵𝘪𝘰𝘯 𝘪𝘯 𝘕𝘦𝘸𝘴 𝘙𝘦𝘤𝘰𝘮𝘮𝘦𝘯𝘥𝘢𝘵𝘪𝘰𝘯𝘴 𝘸𝘪𝘵𝘩 𝘵𝘩𝘦 𝘊𝘭𝘶𝘴𝘵𝘦𝘳𝘪𝘯𝘨 𝘰𝘧 𝘕𝘦𝘸𝘴 𝘚𝘵𝘰𝘳𝘺 𝘊𝘩𝘢𝘪𝘯𝘴” is accepted at NORMalize 2023, co-located with RecSys 2023! 🎉 This work, a summary here, is based on first-author Alessandra Polimeno’s excellent master thesis work, supervised by me and co-authors Sanne Vrijenhoek and Antske Fokkens. I will present it on Sept 19 in the NORMalize workshop.}

_{12 June - 1 Sept 2023 “Computational Linguist & AI” summer intern in the “Content & Creator AI Linguist” team at LinkedIn in Dublin, Ireland! 🍀}

_{24 - 30 May 2023 At ICA23: Presented a paper on “Argument Mining to Analyze Reasons for (a Lack of) Trust in Sustainable Initiatives” - written with Ana Isabel Lopes (communication science) and student Alessandra Polimeno (CLTL). Also participated in the pre-conference hackathon with a case on few-shot detection of hypocrisy accusations in online sustainability debates, in an interdisciplinary team of PhDs.}

_{4 May 2023 Participated in the CLTL hackathon on generative models - in an excellent team with fellow PhDs Urja Khurana and Jonathan Kamp and MA student Swarupa Hardikar. We investigated a case study involving consistency in multi-lingual and multi-cultural settings, including computational subjective argumentation.}

_{24 Feb 2023 Invited for a short talk, “Cross-topic Stances and How to Find Them (in the News)”, at the University of Groningen, with also talks by fellow Argument Miners from CLTL!}

_{6 to 8 Feb 2023 Participated in the 2023 HPLT winter school on Large-Scale Language Modeling in Norway, and wrote a blog post about my personal highlights.}

_{21 - 25 November 2022 Invited participant to Lorentz workshop week “Diverse News Recommenders: from concept to implementation”, where I gave a talk on my own and master student Alessandra Polimeno’s work, and participated in implementing NLP for diversity in news recommendation.}

_{12 August 2022. We (a team of CLTL PhDs) won 🥇 the shared task of the 9th Argument Mining Workshop at COLING 2022. We combined inter-task-training on NLI, contrastive learning, & prompting for Argument Novelty and Validity prediction. Our paper was also published in the Proceedings.}