Show threadUKP LabMay 21
He joins us from MBZUAI (Mohamed bin Zayed University of Artificial Intelligence), where he completed his MSc in Natural Language Processing #NLP. His previous work spans 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝘁𝗲𝘅𝘁 𝗴𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻 with large language models, 𝗩𝗶𝗲𝘁𝗻𝗮𝗺𝗲𝘀𝗲 𝗡𝗟𝗣, and 𝗵𝗮𝘁𝗲 𝘀𝗽𝗲𝗲𝗰𝗵 𝗱𝗲𝘁𝗲𝗰𝘁𝗶𝗼𝗻, with publications at venues like #EACL, #NAACL, and #semeval
Osma SuominenMay 9, 2025

In the first #LLMs4Subjects challenge at the SemEval-2025 workshop, our #Annif team did very well!

The challenge was to generate good quality subject indexing for bibliographic records in German & English using LLMs. We used LLMs for data preprocessing (translation & synthetic data) and Annif as the main suggestion engine. We ranked 1st and 2nd in quantitative and 4th in qualitative evaluations out of 14 teams!

More info & preprints: https://groups.google.com/g/annif-users/c/b8kVy6XSzB4/m/JE6xBzSuEgAJ

#subjectIndexing #AI #LLM #SemEval

Annif awarded at the LLMs4Subjects challenge

Kilian EvangOct 28, 2022
Clickbait spoiling – a much needed new #nlproc task!
---
RT @phschaer
It was a pleasure to welcome @maik_froebe from @webis_de at @th_koeln. Cool overview on #SemEval Clickbait Spoiling https://pan.webis.de/semeval23/pan23-web/clickbait-challenge.html
https://twitter.com/phschaer/status/1584439478789079040
Clickbait Challenge at SemEval 2023 - Clickbait Spoiling

Clickbait posts link to web pages and advertise their content by arousing curiosity instead of providing informative summaries. Clickbait spoiling aims at generating short texts that satisfy the curiosity induced by a clickbait post.