Leonardo Neves

Scientist, Manager, ML instructor, Flamengo fan and musician wannabe.

Currently Leading the Computational Social Science team at Snap Research

[Jose Camacho Collados @CamachoCollados]

Check out TweetNLP if you haven't already done so!

[Jose Camacho Collados @CamachoCollados]

The submission deadline for EvoNLP is quickly approaching! #emnlp2022 #NLProc #evonlp2022

📆1⃣October 10th for regular submissions
📆2⃣October 25th for the ARR track

[Leonardo Neves @lnevespapers]

@bhuwandhingra Congratulations, my friend!

[Francesco Barbieri @fvancesco]

Happy to see more works in this direction! 👏👏👏

Twitter LMs models updated every three months since Jan 2020 can be found online at:
https://t.co/YzronurIXX
(last model from Jun2022!)

[Leonardo Neves @lnevespapers]

@BrihiJ @Snap @sarahookr Loved meeting you there!

[Dan Lou @danielbloureiro]

We're expecting this will be a great venue to share and discuss early works on temporal adaptation of LMs, among other topics. Submission deadline on October 10! #NLProc #EMNLP2022

[Ever Evolving NLP @Evo_NLP]

The first call for papers of the Workshop on "Ever Evolving NLP" is out! https://t.co/9NLtuq9oVF

If you are interested in discussing the challenges posed by the dynamic nature of language, this is your workshop! #NLProc #EMNLP2022

[Francesco Fabbri @Fra_Fabbri]

@ahhhhhjulie rocking at #ICWSM2022 presenting "Sunshine with a Chance of Smiles: How Does Weather Impact Sentiment on Social Media?" showing how weather and location features can improve the performances in sentiment analysis tasks.

@icwsm

[Cardiff NLP @Cardiff_NLP]

So it happened, our CardiffNLP sentiment analysis model is currently the most popular model in @huggingface! 🔥

Downloaded over 15 million times last month. Kudos to our group members @CamachoCollados and @espinosaanke, and our colleagues at Snap @fvancesco and @lnevespapers 👏

[Leonardo Neves @lnevespapers]

Time is an important aspect when understanding society and how people communicate. We have recently released timelms, an initiative to address the issue of language changes and model degradation in NLP models for social platforms. Check it out!

[Jose Camacho Collados @CamachoCollados]

We have a problem if our most popular NLP models (BERT, RoBERTa, etc.) don’t even know what COVID-19 is.

Solution? We launch our initiative TimeLMs, with the commitment to release a new language model trained on recent data every three months! 👇

https://t.co/kYfBMbk3EX

[Jose Camacho Collados @CamachoCollados]

This year we are planning to organise 🔥EvoNLP🔥 The 1st Workshop on Ever Evolving NLP.

EvoNLP will be an industry and academic forum to discuss the challenges posed by the dynamic nature of language, especially in the current #NLProc paradigm dominated by language models.

[Leonardo Neves @lnevespapers]

@kous2v @munmun10 @gabowd @thomasploetz @emrek @Gmark6 Congratulations!!

[Francesco Barbieri @fvancesco]

Snap Research is hiring interns for 2022!

The Computational Social Science team under Research is looking for interns in these research areas:
- NLP
- Computational Social Science
- User behavior modeling
- Graph representation learning and mining

https://t.co/XvsOXHWI1N

[Jose Camacho Collados @CamachoCollados]

Glad this (unusual) #NLProc paper is out, thanks to @asahiushio and F. Liberatore #emnlp2021

📊A Quantitative Analysis of Statistical and Graph-Based Term Weighting Schemes for Keyword Extraction: https://t.co/dinwOXthDF

💾Keyword extraction library: https://t.co/ABuPlcaraw

[Leonardo Neves @lnevespapers]

@CamachoCollados super interesting work, I love it! congrats!

[Farhan Asif Chowdhury @farhanasifchy]

@kous2v @nickmvincent @maartenwbos @lnevespapers Looking forward to present our work on improving user behavior prediction on social platforms by better leveraging individual user's regularities at ICWSM'21. This work was done during my summer'20 internship at Snap w
@kous2v @nickmvincent @lnevespapers @maartenwbos.
#ICWSM21

[Intel® Tiber™ AI Studio @cnvrg_io]

New #mlcon Speaker Announcement! Welcome to the stage
@lnevespapers from @snapchat Maria Lee from @roche
@DrScottClark GM of @SigOpt Rushen Patel from @marksandspencer @likeRyanwithaT from @Zulily
Mojtaba Farmanbar, Ph.D. from @ING

See more speakers: https://t.co/xfp7OHFRuF

[Leonardo Neves @lnevespapers]

I had the honor of being the Keynote Speaker for the SDPRA Workshop yesterday - check my blog for a summary of the talk and links to all the presentations in the workshop. https://t.co/L57QfJAtdM

[arXiv cs.CL 自動翻訳 @arXiv_cs_CL_ja]

XLM-T: A Multilingual Language Model Toolkit for Twitter https://t.co/J5zcTWtpyV
Comment: Submitted to ACL demo. Code and data available at
https://t.co/pNZy6VsbUO
言語モデルは現在のNLPに遍在しており、その多言語能力は最近かなりの注目を集めています。ただ

[Leonardo Neves @lnevespapers]

This thread can be read here: https://t.co/uUh1lId1PJ

[Leonardo Neves @lnevespapers]

Reviewer #2 strikes again https://t.co/uUh1lId1PJ

[Leonardo Neves @lnevespapers]

Do your job, but do both parts of it and help the authors grow where you can.

Related to this, I will be giving a keynote @ SDPRA 2021 where I will touch on these topics and make a case for simplicity in research. Please join me in May for that, if you are interested!