David

.@anaezquerro_ presenting our work on graph parsing as tagging at #EMNLP2025 !!!

Today, @anaezquerro_ will be presenting our paper 'Hierarchical Bracketing Encodings for Dependency Parsing as Tagging' at #ACL2025NLP 🇦🇹 Room 1.86 (Session from 14:00 to 15:30). Paper: aclanthology.org/2025.acl-long.…

Happy to share that our paper 'Document-level event extraction from Italian crime news using minimal data' has been published: sciencedirect.com/science/articl… #nlp #nlproc

📓Baseline Notebooks: - Fine-tuning BERT-based models - NCRF++ models via GitHub 💡Check the repo! github.com/mmaguero/guara… 📖Check out the publication here: - digibug.ugr.es/handle/10481/9… #NLP #AffectiveComputing #LowResourceLanguages #Guarani #Jopara #SentimentAnalysis #AIForAll

Highlights: 🧵Corpora: - Emotion Recognition - Humor Detection - Offensive Language Identification 💻Base Models for Fine-Tuning (trained on Guarani Wiki): - From scratch: BERT-based tiny, small, base and large models - Continuously pre-trained models: m-BERT and BETO

🚀 Multidimensional Affective Analysis for Guarani/Jopara! 🌎 This project explored affective computing for low-resource languages, focusing on emotion recognition, humor detection, and offensive language identification in Guarani and Jopara (a code-switching mix with Spanish).

Today at 14:00, we'll be presenting "Dependency Graph Parsing as Sequence Labeling," by @anaezquerro_, @eusondavid, and @carlosgr_nlp at #EMNLP2024! Join us at the poster session in Riverfront Hall. 🔗 Paper: aclanthology.org/2024.emnlp-mai…

🌴 On our way to Miami for #EMNLP2024 to present two papers! 📄 - Revisiting Supertagging for Faster HPSG Parsing (@OlgaZamaraeva, @carlosgr_nlp) - Dependency Graph Parsing as Sequence Labeling (w/ @anaezquerro_, @eusondavid, @carlosgr_nlp)

@mmbollmann @mdlhx Interesting. In Spain we don't, but I like the idea.

Tfw when resolving a student's name on a handwritten exam is not straightforward.

📰Xa dispoñible no #RUC e #Zenodo o traballo do grupo GI-LYS e da @FIC_UDC: "From Partial to Strictly Incremental Constituent Parsing" aclanthology.org/2024.eacl-shor… hdl.handle.net/2183/36570 zenodo.org/records/113654…

📰Xa dispoñible no #RUC e #Zenodo o traballo do grupo @lysresearch e da @FIC_UDC: "4 and 7-bit Labeling for Projective and Non-Projective Dependency Trees" aclanthology.org/2023.emnlp-mai… hdl.handle.net/2183/36571 zenodo.org/records/113657…

Declined a reviewing invitation and the portal says this 🤔

@ragerri @JulioGonzalo1 @carlosgr_nlp @eagirre @artetxem Exacto. Yo les he escrito ya tambien, y ojalá que lo haga mucha gente más. Esperemos que entren en razón y se consideren como aportaciones ordinarias. No entiendo este movimiento. Lo contrario es un paso atrás en la buena dirección que por fin se había tomado en los últimos años.

@JulioGonzalo1 @carlosgr_nlp @eagirre @artetxem Hay que movilizar a los investigadores de nuestros respectivos centros ya que esto afecta a sexenios pero tb acreditaciones para plazas, etc. Es fundamental remarcar que este criterio volvería a ningunear a los principales foros de publicaciones en IA...

Por primera vez ANECA nos pregunta sobre un borrador de criterios de evaluación para los sexenios de investigación. ¡No desaprovechemos la oportunidad!

@jparapar @anxo6ila @KexinWang @NehaWarikoo @IGurevych ¡Felicidades!

Happy to see @anxo6ila work accepted at #EMNLP2023! In this work with @kexinwang @NehaWarikoo and @IGurevych we present an efficient semantic pipeline for depression severity estimation.

To me, at least in theory, reasons to reject should be something stronger than weaknesses, but in practice, when it comes to reviewing, I see many choose to copy&paste from one field to the other, and many times both fields contain overlapping information.

Finding that many reviewers for #eacl2023, like myself often when reviewing, feel that "strengths/weaknesses" and "reason to accept/reasons to reject" are roughly equivalent. Any views on how these two are important and different?

Cheers mates, for a glorious Wednesday

If you're interested in making NLP/AI equally useful for all of us, and perhaps more responsible, here's a chance to join CoAStaL/UCPH [coastalcph.github.io] as a PhD student: ddsa.dk/event2/ddsa-ph…

Any volunteers out here willing to do an emergency review for EACL 2023 - Multidisciplinary and NLP Applications track?

@kroscoo I do it in my undergraduate courses. I think it gives a nice intuition into the need of more complex models.

Are count-based language models still important paedagogically? NLP professor acquaintances: do you still cover simple n-gram language models in your NLP courses?