Jose Camacho-Collados
@CamachoCollados
Followers
1,908
Following
967
Media
201
Statuses
1,018
Professor @cardiffuni ( @Cardiff_NLP ). Natural Language Processing researcher and chess International Master ♟️
Cardiff, Wales, UK
Joined June 2011
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
ANILPIN BLOSSOM AROMATIC
• 868615 Tweets
Chelsea
• 441901 Tweets
Alain Delon
• 313012 Tweets
Enzo
• 131728 Tweets
Sanchez
• 118487 Tweets
Haaland
• 103990 Tweets
#CHEMCI
• 84368 Tweets
Palmer
• 61637 Tweets
Mallorca
• 57324 Tweets
Acuña
• 48698 Tweets
Kovacic
• 44785 Tweets
#bbnaija2024
• 43495 Tweets
Lavia
• 42642 Tweets
Rojo
• 39910 Tweets
Rodrygo
• 33318 Tweets
Caicedo
• 30679 Tweets
Nkunku
• 29709 Tweets
#BJKvANT
• 25030 Tweets
Ashley
• 25029 Tweets
Anselmino
• 22567 Tweets
Conte
• 21740 Tweets
Anthony Taylor
• 15496 Tweets
Ebuka
• 13409 Tweets
Gimenez
• 12686 Tweets
#انقذوا_النصر_من_صناع_قراره
• 11712 Tweets
Rayan
• 10229 Tweets
Merentiel
• 10003 Tweets
Given the current situation,
@tpilehvar
and I have decided to openly release the first draft of our book “Embeddings in Natural Language Processing”. We also thank
@MorganClaypool
for agreeing to this early draft release.
Link:
21
322
971
📢 Today we officially launch TweetNLP, an all-round NLP platform for social media. From sentiment analysis to emoji prediction and more 🔥🔥
✔️ TweetNLP includes a Python API, a demo and tutorials. Useful for developers and researchers alike. Want to know more? 🧵
9
93
394
I've just been told that our Twitter-based sentiment analysis model is the third most downloaded model in
@huggingface
, just behind the original BERT and GPT. It has been downloaded over 9 million times last month.
That's just INSANE!!!🤯🤯🤯🤯
5
22
251
We release 🔥 XLM-T 🔥 a multilingual language model trained on Twitter for 30+ languages (~200M tweets)
#NLProc
What else?
🌍 Unified sentiment analysis dataset for 8 languages
🤗
@huggingface
models and starter code
3
58
236
Yesterday I was blocked and insulted by Vladimir Kramnik on this platform. I was also accused of not being IQ friendly, whatever that means.
Please judge the conversation below by yourselves to see if this was warranted 👇
33
10
224
Train and test NER transformer models with a single line of code! Check out
@asahiushio
's open-source T-NER library for named entity recognition:
Bonus:
- Pre-trained (xlingual) NER models in the
@huggingface
hub 🤗
- Nine datasets 📚
- A cool web app 🖥️
3
50
220
After many months working on it, including the birth of our respective first sons in-between :),
@tpilehvar
and I are happy to finally share our survey "From Word to Sense Embeddings: A Survey on Vector Representations of Meaning"
Link: 👈
4
48
165
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! 👇
6
26
161
Congratulations to Adriana Cerezo for winning the silver medal in Taekwondo, first medal for Spain! 🏆
#Tokyo2020
Btw, fellow Korean NLPers, what do you think of the Korean sentence on her black belt? "기차 하드, 꿈 큰" -> "Train (🚂) hard, dream big"?😅 cc:
@kchonyc
4
115
119
I've been promoted to (Full?) Professor, effective from August 1st. I feel that I haven't fully digested the news yet...
This wouldn't have been possible without the amazing team at
@Cardiff_NLP
,
@CompScienceCU
and all the great collaborators over the years. Thank you all! 🤗
15
5
119
Happy to announce that I’ve been promoted to Senior Lecturer!😊🥳
Thank you to everyone who has supported me during the years. And of course, this wouldn’t have been possible without the amazing human team at
@Cardiff_NLP
and
@CompScienceCU
! 🙏
11
2
94
Our
#AAAI2020
preprint “Inducing Relational Knowledge from BERT” (w/ Z. Bouraoui and S. Schockaert) is now out:
In this paper, our aim is to understand to what extent pre-trained language models like BERT capture relational knowledge. A short thread 👇
3
13
77
✅Is there a problem? Yes
✅Does something need to change? Yes
✖️Dodgy methodology and spreading unfounded accusations is the way to fix it? Certainly not, and it actually damages the cause.
/END
2
0
74
If you need high-quality pre-trained monolingual and cross-lingual embeddings for social media (in different languages), you can download them here:
Our preprint “Learning Cross-lingual Embeddings from Twitter via Distant Supervision” is now out!
We compiled large Twitter corpora for five languages (English, Spanish, Italian, German, Farsi) to learn unsupervised cross-lingual embeddings. Thread 👇
2
11
26
2
18
70
Our "Embeddings in NLP" book has been officially published!😱
Still cannot believe it - what a journey with
@tpilehvar
to get to this point.
This version has significantly improved since the early draft release - thanks to everyone that made it possible!
Given the current situation,
@tpilehvar
and I have decided to openly release the first draft of our book “Embeddings in Natural Language Processing”. We also thank
@MorganClaypool
for agreeing to this early draft release.
Link:
21
322
971
2
11
70
A year and a half in the making but is finally out. An “old-fashioned” journal article with lots of analysis (50 pages!) and an overview of language models for word sense disambiguation. How well does BERT capture lexical ambiguity? Check it out!
Paper 👉
3
20
61
Officially starting my UKRI Future Leaders fellowship today! 😱 Super excited to start this journey of 4+ years to work on
#NLProc
, language understanding and beyond.
#UKRIFLF
@Cardiff_NLP
▶️
6
3
58
As a mathematician and AI researcher, my eyes just hurt every time he uses the words “mathematics” or “statistics”.
There is no robust mathematical methodology behind his analyses and he keeps mixing some data that does make some sense, with some personal crusades with no value.
1
4
55
Super excited and honoured to be awarded a UKRI Future Leaders Fellowship to work on language understanding. Looking forward to getting started soon!
#UKRIFLF
Congratulation to Jose Camacho Collados
@CamachoCollados
Cardiff University
@cardiffuni
one of our new Round Four
@UKRI_News
Future Leaders Fellows
#UKRIFLF
1
7
34
10
3
56
Today
@asahiushio
successfully defended his PhD thesis.
It's been an immense pleasure to accompany you during this journey. I'm still amazed of what you've achieved in this short period and excited to see what comes next!
#proudadvisor
Congratulations, Dr Ushio! 👏👨🎓
2
6
53
A few days ago I wrote a short post on the application and general implicit benefits of using neural networks and word embeddings in NLP (especially targeted to newcomers)
#NLProc
0
7
48
It's amazing to see
@Cardiff_NLP
as the 7th ranked organisation in terms of open ML contributions in
@huggingface
, with >11M monthly downloads 🚀
Just behind Google and 1st university lab in the world! 🤯
Check out this
@huggingface
organizations leaderboard created by a community member!🤯
-
@StabilityAI
is the org with the most likes (15k)
-
@MetaAI
is among the orgs with most models (700!)
-
@runwayml
has the highest avg of downloads per model
8
31
151
1
8
50
Winner of the South Wales Chess International 2019! 🏆
After a long period of inactivity, it's nice to see that I can still play some chess! 😅 Winning arguably one of the strongest tournaments of my career.
6
2
45
Tomorrow at
#emnlp2022
we are presenting our TweetNLP platform (2pm, Demo session).
We have prepared some important improvements with respect to the previous version, check below! 👇
1
12
46
I pointed out issues with his methodology, as others have done. Basically, if you torture your data long enough, it will confess whatever you want. I tried to give a similar example to make my point, which he didn't take well...
This isn't a statistical anomaly, but it's a perfect example of another well-known phenomenon, known as "confirmation bias".
Here you can read a bit more about what confirmation bias is and how to overcome it 👇
2
0
21
1
1
45
These two can be simultaneously true:
1) ChatGPT is, by far, the most impressive general-purpose chatbot we have ever seen.
2) There are still many exciting challenges in NLP that are not going to be solved by simply scaling up language models (even for dialogue systems!)
2
1
44
After recent discussions in Twitter, it seems increasingly clear that word analogies weren't as promising as initially anticipated. Check our
#ACL2019
paper () to learn how to really encode relational knowledge into word embeddings in an unsupervised way!
2
8
42
Our paper “Learning Cross-lingual Word Embeddings from Twitter via Distant Supervision” was accepted to
#ICWSM2020
(
@icwsm
)!
Paper 👉
Data 👉
Data include monolingual and cross-lingual embeddings from Twitter. More info below:
Our preprint “Learning Cross-lingual Embeddings from Twitter via Distant Supervision” is now out!
We compiled large Twitter corpora for five languages (English, Spanish, Italian, German, Farsi) to learn unsupervised cross-lingual embeddings. Thread 👇
2
11
26
1
14
41
The Cardiff NLP group is awesome, super proud of the team and congrats to everyone! ☺️✨
The team is very diverse which is reflected on the variety of research topics. Stay tuned for the preprints and also looking forward to reading all the papers myself!
#emnlp2023
📢 Check out the accepted EMNLP papers from our Cardiff NLP group!
Stay tuned for the camera-ready preprints 🔜
#emnlp2023
#cardiffnlp
#NLProc
1
7
53
2
5
40
1
14
39
More relational papers coming in... Pleased to announce that "Relational Word Embeddings" got accepted to
#acl2019nlp
@ACL2019_Italy
! 😀 With
@luisanke
and Steven Schockaert. PyTorch code+preprint coming very soon.
#NLProc
0
2
38
One of the most promising researchers in Spain over the past few years, with an impressive CV and publication record. He could develop his career anywhere in the world, being highly valued... except in Spain, where his CV is considered mediocre.
@CientificoenEsp
2018 was a productive year! I am 3rd in this ranking with 5 first author papers in top ML/NLP venues
Unfortunately, the Spanish academic system is broken, and these publications may count close to zero if I decide to develop my career here. Thread 👇🏼
23
104
315
0
22
37
We all know about Question Answering and its cool applications, but what about Question *Generation*? Check it out!
#emnlp2022
📜 Walk-through of QG methods, benchmarks and evaluation:
💻 Demo in multiple languages!
1
11
37
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.
1
18
39
Because of this, I disagree with
@LevAronian
's take. I agree with the sentiment, but the means are important.
Guys like Kramnik,who question everything and would be heroes like Frank Serpico in the age free speech,are now perceived like public enemies because people want the comfort of not knowing.
I may agree or disagree with his opinions but Vlad is important and i value his efforts.
104
45
795
3
1
37
I'm honoured to receive this award. As both AI researcher and chess player, it feels surreal to receive it alongside the Deep Blue paper (AIJ 2023 Classic Paper Award!) 🤯
The
@Kasparov63
vs Deep Blue match was my first approximation to AI and why I got interested in the field.
Congratulations to José Camacho-Collados, Mohammad Taher Pilehvar, and Roberto Navigli for winning the AIJ 2023 Prominent Paper Award!
Read the winning paper here:
#ArtificialIntelligence
#AI
0
7
23
2
5
37
Chess is so unpredictable. Less than 24 hours after my worst blitz tournament in years (dropping 77 FIDE points!), I managed to get a very decent 6th at the Spanish Rapid Championship with >60 titled players.
0
2
36
The fact that he suspects of players after five (!) moves (see this video after minute 23:18 as one of the many examples) is the confirmation of a paranoia that has gone too far.
2
1
36
The
#NLProc
experimental landscape in social media is quite fragmented...not anymore!
We introduce 🔥TweetEval🔥: a unified benchmark for tweet classification with 7 popular tasks.
#emnlp2020
(Findings)
Paper 👉
Benchmark 👉
1
8
35
Now travelling to the beautiful city of Donostia-San Sebastián for the Global WordNet Conference. Honoured to have been invited as one of the keynote speakers!
#gwc2023
2
5
35
🚂 Off to Cambridge for a short visit and talk tomorrow at
@CambridgeLTL
!
Of course, this wouldn't be a proper academic trip if I weren't working on the presentation last-minute in the train 😂
2
2
34
In our latest
#EMNLP2023
paper, we present a systematic analysis of the relevance of different factors (model size, training objective, tokenization, training data) in LM social biases and downstream performance.
📒 It includes a detailed survey of 39 LMs of diverse nature!
Wonder what underlying factors influence the social biases in masked language models? In our
#EMNLP2023
paper with
@CamachoCollados
and
@Bollegala
, we show that model size, tokenization and training objectives are the three most important factors.
📜:
0
5
31
1
5
32
Are unambiguous words useful for Word Sense Disambiguation? Surprisingly yes! Check out our latest work with
@danielbloureiro
to find out why and how.
Paper 👉
(Note: for full transparency, we've also made our
#acl2020nlp
reviews available)
#NLProc
1/2
2
5
32
Two intense days at the
#cardiffnlpworkshop
with exciting talks, tutorials, panels, poster presentations...
Glad to meet so many people, old friends and new ones. Thank you all for coming to Cardiff and making it possible! 🤗
1
9
30
Excited to announce that I've been appointed as lecturer at
@cardiffuni
! Officially starting next month.
I'm happy to continue my career here in Cardiff, and to be able to contribute as part of a growing
#NLProc
team at
@CompScienceCU
.
1
1
32
Our poster presentations today at
@NLPerspectivWS
!
Many thanks to the organisers for a great workshop and
@barbara_plank
for an enlightening keynote talk!
0
1
32
Two papers rejected at
#naacl2019
with 5-5-3 and 5-4-3 scores. Two reviews contained factually wrong statements. Not blaming the reviewers as the reviews were mostly constructive, but in hindsight I'm wondering whether the rebuttal would have helped in this case...
2
1
31
I was just very annoyed. The way he has been approaching all sorts of players and accusing them of cheating without evidence is just going too far. Especially when I know that he has explicitly accused some people that have not cheated at all.
2
1
31
In the following piece published in
@chess24com
, I provide a statistical overview of the current gender gap in chess across countries, including experiments about the relation between low participation and the presence of women at the top level. (1/7)
1
7
29
We are pleased to release the full XL-WiC benchmark! 🚀
🌎13 languages: Bulgarian, Chinese, Croatian, Danish, Dutch, English, Estonian, Farsi, French, German, Italian, Japanese and Korean.
✅ Evaluation scorer and gold labels for all languages.
👉
2
9
31
Today presenting
@comuni_carla
's work on mansplaining.
Lot of interest in the topic and great discussions!
#LRECCOLING2024
@Cardiff_NLP
1
3
30
Proud to have been selected to represent Wales at the Chess Olympiad! 🏴
(apologies for the spoiler, but check out the video to discover the full Olympiad Welsh open team!)
0
3
5
3
2
30
Excited to start a challenging project to detect meaning shift in social media with
@espinosaanke
and
@danielbloureiro
. Thank you
@Snap
for the research grant and funding this project - looking forward to getting started with the collaboration!
0
4
30
Thanks
@stanfordnlp
for this extremely useful and high-quality
#NLProc
library. Available for 53 languages! 🙏
Out now: our new Python
#NLProc
package. StanfordNLP provides native, neural (PyTorch) tokenization, POS tagging and dependency parsing for 53 languages based on UD v2—and a Python CoreNLP interface. PyPI: – pip install stanfordnlp
13
756
2K
1
14
30
Do you need to model relations such as "Tokyo-Japan" or even "dog-bark"? We openly release 🔥RelBERT🔥 a language model specifically fine-tuned to produce high-quality word pair embeddings
#NLProc
#emnlp2021
🤖Model+code:
📄Paper:
1
9
29
“Language Models for Text Classification: Is In-Context Learning Enough?”
@AleksandraEdwa3
now presenting at the poster session, come join us! 👋🏼
#LRECCOLING2024
0
2
29
After living in Cardiff for the last five years, I'm proud to call Wales home. From today, Wales will also be my new home chess-wise.
Diolch, Cymru! 🏴
1
0
29
#COLING2020
call for papers is now out!
@tpilehvar
and I will give a tutorial on "Embeddings in
#NLProc
", a historical journey from word embeddings (Word2Vec, GloVe, etc.) to the more recent contextualized variants of Muppetland (ELMo, BERT, etc.).
See you all in Barcelona!
COLING 2020 first call for papers (main conference) is up on the website!
1
58
93
0
10
28
Our preprint “Learning Cross-lingual Embeddings from Twitter via Distant Supervision” is now out!
We compiled large Twitter corpora for five languages (English, Spanish, Italian, German, Farsi) to learn unsupervised cross-lingual embeddings. Thread 👇
2
11
26
Had a great time in Turin! Thank you to everyone who made it possible, especially to
@Ginger_in_AI
,
@vivi_patti
and all the organisation team.
I particularly liked the poster sessions. They didn't quite get the weather right though, something to improve for the next one! 😅
That's a wrap! 😢
#LRECCOLING2024
has come to an end after an active participation from our group.
See you next year at
#COLING2025
and a year later in
#LREC2026
! 👋
0
2
18
1
3
28
Good morning
#acl2019nlp
! Come check our poster today (10:30-12:10) if you are interested in knowing how to encode relational knowledge into word embeddings without the need of external resources.
@luisanke
, Steven and I will be happy to walk you through our work!
After recent discussions in Twitter, it seems increasingly clear that word analogies weren't as promising as initially anticipated. Check our
#ACL2019
paper () to learn how to really encode relational knowledge into word embeddings in an unsupervised way!
2
8
42
1
5
27
“Is AlphaZero really a scientific breakthrough in AI?” //TL;DR: Hard to tell…
1
15
25
Two papers accepted to
#AAAI2020
as oral presentations! "Inducing Relational Knowledge from BERT" (with Zied Bouraoui and Steven Schockaert) and "Modelling Semantic Categories using Conceptual Neighborhood" (same, plus
@espinosaanke
). Preprints coming soon.
2
6
28
Last poster of the day: Databench, a new benchmark on Question Answering over Tabular Data! Led by
@_jorses
and co-presented with
@genimarca
and
@laurenalopez
#LRECCOLING2024
🔜Databench will feature as a task in SemEval 2025, stay tuned!
0
3
25
And from today, I'm the new Welsh champion! 🏴🏆
It's been a roller coaster of a tournament, but I'm proud to end undefeated after saving a few very complicated games.
@WelshChessUnion
After living in Cardiff for the last five years, I'm proud to call Wales home. From today, Wales will also be my new home chess-wise.
Diolch, Cymru! 🏴
1
0
29
2
2
26
From today,
@tpilehvar
and I will be teaching a course on "Embeddings in Natural Language Processing" at
#ESSLLI2021
@ESSLLI_official
.
We will update the content of the course (including slides and a couple of Jupyter Notebooks) as we go along.
1
2
25
We have created a website with information about our recent "Embeddings in NLP" book, including how to get it and a short video overview (presented as a tutorial at
#COLING2020
).
@tpilehvar
#NLProc
0
6
25
Words can be recovered from word vector averages, as shown by
@prfsanjeevarora
et al. In our paper we show that word order can also be recovered by simply adding bigrams into the equation. At the same time, we provide some evidence on why LSTM autoencoders work in practice.
1
1
24
I just wish he realizes the damage he is causing and goes back to use reason and data in a sensible way.
His input and insights would surely be super valuable! But first, he must stop these personal crusades and indiscriminate accusations that go nowhere.
1
0
24
A pleasure to host
@seb_ruder
at
@CompScienceCU
today. He is now giving an engaging talk on "Cross-lingual transfer learning" as part of the DKE Seminar series.
1
1
24
The 3rd Cardiff NLP Workshop is just around the corner, join us! 🔜
The registration period for the 3rd Cardiff NLP Workshop is now open!
📅 Registration deadline: June 5th or until all places are filled.
Check out the amazing list of speakers and don't miss it!
1
14
22
0
10
22
✈️ Flying to Abu Dhabi tomorrow for
#emnlp2022
, my first in-person conference since 2019!
Looking forward to lots of
#NLProc
, and especially to meeting three-dimensional new and old friends and colleagues!
1
0
23
Just back home after an intense day at
#EurNLP2019
. Thought-provoking talks, good atmosphere and nice opportunity to meet old (and new!) friends. Thanks to all the organizers and presenters!
0
0
22
At
@CompScienceCU
we run a "Data and Knowledge Engineering" seminar every Monday. Check out the amazing list of speakers for the Spring semester, with a great mix of internal and external speakers from both industry and academia!
1
2
23
And here a picture with six members of the team presenting the TweetNLP demo at
#emnlp2022
!
Tomorrow at
#emnlp2022
we are presenting our TweetNLP platform (2pm, Demo session).
We have prepared some important improvements with respect to the previous version, check below! 👇
1
12
46
0
4
22
All the content (slides, outline and references) of our
#NAACL2018
Tutorial on "The Interplay between Lexical Resources and
#NLProc
" is available here:
@luisanke
@tpilehvar
0
10
22
And
#evonlp2022
is over! Thank you all for participating in the workshop.
Hope the workshop sparked a discussion on the importance of going beyond static datasets/models by taking into account the dynamic nature of language (time, language change, etc.) in
#NLProc
research.
1
4
22
Happy to see our young group growing! And one interesting trend: 86% (6/7) of all
@Cardiff_NLP
first authors are women! 👩🔬👩🔬👩🔬
We’ve now extended our list of accepted papers to be presented in the next couple of months with
#coling2020
,
#ijcai2020
and workshops. Preprints coming soon!
#emnlp2020
#conll2020
#aacl2020
0
1
13
1
5
22
Do LLMs understand the subtle language involved in mansplaining situations? 🤔
Well, not really... LLMs can usually detect mansplaining if explicitly asked, but not otherwise.
👩🔬Work led by
@comuni_carla
📜Corpus with 886 mansplaining stories
Paper:
1
2
22
Thanks for the invitation! I'll be presenting an overview on embedding techniques for Natural Language Processing, including some of our latest work in this area.
If that doesn't convince you to attend, this will surely will... there'll be free pizza!🍕😂
1
3
20
Some take-away messages from the SemEval task on Hypernym Discovery: Numbers don't lie, it is a *very* difficult task (room for improvement!); Hearst patterns still relevant; most frequent baseline hard to beat in focused domains.
Paper here:
1
5
20
This isn't a statistical anomaly, but it's a perfect example of another well-known phenomenon, known as "confirmation bias".
Here you can read a bit more about what confirmation bias is and how to overcome it 👇
2
0
21
“Takeaways from a systematic study of 75K models on Hugging Face”, presented by
@nazneenrajani
now at
#evonlp2022
#emnlp2022
1
4
21
📢Hiring!📢 Postdoc position for up to three years at
@Cardiff_NLP
(
@CompScienceCU
) to work on an exciting project on language understanding.
Deadline: Dec 14th. Start date: from Apr 2021 (flexible)
Call for applications 👉
Please RT!
#NLProc
#UKRIFLF
0
13
21
The EvoNLP workshop is under way! Now
@eunsolc
on “Knowledge-rich NLP models in a dynamic real world”
#evonlp2022
#emnlp2022
@Evo_NLP
0
1
21
Everybody around me knows how vocal I am to increase anti-cheating measures on all sorts of tournaments (even OTB, we need better measures). However, the way he is going about all this is just wrong, and thankfully most people can see that by now.
1
0
20
Check out the thread about SuperTweetEval, our newest benchmark for NLP research on social media!
#NLProc
#EMNLP2023
🤗 Data and models publicly available!
🧵 Introducing "SuperTweetEval: A Challenging, Unified and Heterogeneous Benchmark for Social Media NLP Research"
1
10
23
0
6
20
Just arrived in Daejeon🇰🇷, where I'll spend the next six weeks visiting
@aliceoh
's group at KAIST.
Looking forward to it and very excited to get to know more about the Korean NLP landscape!
0
1
20
On my way to Torino ✈️🇮🇹 for
#LRECCOLING2024
!
Ping me up if you want to meet or chat during the conference!
Otherwise, you can also find me in the following presentations 🧵
1
1
20
Great work by Yerai Doval (w/
@luisanke
S.Schockaert) accepted
@emnlp2018
. Building on the works of
@artetxem
and
@alex_conneau
et al for learning crosslingual embeddings, we show how a final transformation can further approach the monolingual spaces while improving their quality
1
4
20
All our models, including XLM-T and a multilingual sentiment analysis model fine-tuned on 8 languages, already available at the
@huggingface
hub:
🤗 XLM-T:
🤗 Sentiment model:
1
5
19
This is neat!Take a corpus in language A and another one in language B, and you can automatically build a dictionary from A to B
#acl2017nlp
2
3
18
Tomorrow
@espinosaanke
and I will give an overview of the
#NLProc
research and related activities conducted at
@CompScienceCU
#WalesTechWeek
Upcoming event: Natural Language Processing at Cardiff University
#WalesTechWeek
Join
@CompScienceCU
for an overview of some of the main research topics being pursued at Cardiff University in the area of Natural Language Processing (NLP).
>>
0
3
5
0
5
19
@VBkramnik
For the record, this player is GM Vladimir Kramnik.
If it wasn't clear, I'm certainly not suspecting anybody has cheated!
But hope this example can make a point: these are *not* statistical anomalies and examples like this can be easily found for *any* player.
1
2
19
Thesis finally submitted! 😁 ... now off to Seoul ✈️🇰🇷
1
0
18
Our paper "A Latent Variable Model for Learning Distributional Relation Vectors" has been accepted to
#IJCAI2019
! Joint work with
@luisanke
,
@shoaibjameel123
and Steven Schockaert. Camera-ready and code coming soon.
0
3
17
I understand that I was too sarcastic and what I said could have been taken the wrong way. I apologise, I should have presented my argument in a more gentle manner.
3
0
18