finiteautomata commited on
Commit
d21e754
1 Parent(s): d3c1bec

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +85 -7
README.md CHANGED
@@ -9,14 +9,92 @@ dataset_info:
9
  dtype: string
10
  splits:
11
  - name: train
12
- num_bytes: 3358310095
13
- num_examples: 24898932
14
  - name: test
15
- num_bytes: 905774909
16
  num_examples: 6224733
17
- download_size: 2628886986
18
- dataset_size: 4264085004
19
  ---
20
- # Dataset Card for "spanish-tweets-small"
21
 
22
- [More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9
  dtype: string
10
  splits:
11
  - name: train
12
+ num_bytes: 82649695458
13
+ num_examples: 597433111
14
  - name: test
15
+ num_bytes: 892219251
16
  num_examples: 6224733
17
+ download_size: 51737237106
18
+ dataset_size: 83541914709
19
  ---
 
20
 
21
+ # spanish-tweets-small
22
+
23
+ ## A smaller version of spanish-tweets
24
+ ## A corpus of tweets for pretraining embeddings and language models
25
+
26
+
27
+ ## Table of Contents
28
+ - [Table of Contents](#table-of-contents)
29
+ - [Dataset Description](#dataset-description)
30
+ - [Dataset Summary](#dataset-summary)
31
+ - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
32
+ - [Languages](#languages)
33
+ - [Dataset Creation](#dataset-creation)
34
+ - [Curation Rationale](#curation-rationale)
35
+ - [Source Data](#source-data)
36
+ - [Additional Information](#additional-information)
37
+ - [Dataset Curators](#dataset-curators)
38
+ - [Licensing Information](#licensing-information)
39
+ - [Citation Information](#citation-information)
40
+ - [Contributions](#contributions)
41
+
42
+ ## Dataset Description
43
+
44
+ - **Homepage**: https://github.com/pysentimiento/robertuito
45
+ - **Paper**: [RoBERTuito: a pre-trained language model for social media text in Spanish](https://aclanthology.org/2022.lrec-1.785/)
46
+ - **Point of Contact:** jmperez (at) dc.uba.ar
47
+
48
+ ### Dataset Summary
49
+
50
+ A big dataset of (mostly) Spanish tweets for pre-training language models (or other representations).
51
+
52
+ ### Supported Tasks and Leaderboards
53
+
54
+ Language Modeling
55
+
56
+ ### Languages
57
+
58
+ Mostly Spanish, but some Portuguese, English, and other languages.
59
+
60
+ ## Dataset Structure
61
+
62
+
63
+ ### Data Fields
64
+
65
+ - *tweet_id*: id of the tweet
66
+ - *user_id*: id of the user
67
+ - *text*: text from the tweet
68
+
69
+ ## Dataset Creation
70
+
71
+ The full process of data collection is described in the paper. Here we roughly outline the main points:
72
+
73
+ - A Spritzer collection uploaded to Archive.org dating from May 2019 was downloaded
74
+ - From this, we only kept tweets with language metadata equal to Spanish, and mark the users who posted these messages.
75
+ - Then, the tweetline from each of these marked users was downloaded.
76
+
77
+
78
+ This corpus consists of ~30M tweets.
79
+
80
+ Please note that we did not filter tweets from other languages, so you might find English, Portuguese, Catalan and other languages in the dataset (around 7/8% of the tweets are not in Spanish)
81
+
82
+ ### Citation Information
83
+
84
+ ```
85
+ @inproceedings{perez-etal-2022-robertuito,
86
+ title = "{R}o{BERT}uito: a pre-trained language model for social media text in {S}panish",
87
+ author = "P{\'e}rez, Juan Manuel and
88
+ Furman, Dami{\'a}n Ariel and
89
+ Alonso Alemany, Laura and
90
+ Luque, Franco M.",
91
+ booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference",
92
+ month = jun,
93
+ year = "2022",
94
+ address = "Marseille, France",
95
+ publisher = "European Language Resources Association",
96
+ url = "https://aclanthology.org/2022.lrec-1.785",
97
+ pages = "7235--7243",
98
+ abstract = "Since BERT appeared, Transformer language models and transfer learning have become state-of-the-art for natural language processing tasks. Recently, some works geared towards pre-training specially-crafted models for particular domains, such as scientific papers, medical documents, user-generated texts, among others. These domain-specific models have been shown to improve performance significantly in most tasks; however, for languages other than English, such models are not widely available. In this work, we present RoBERTuito, a pre-trained language model for user-generated text in Spanish, trained on over 500 million tweets. Experiments on a benchmark of tasks involving user-generated text showed that RoBERTuito outperformed other pre-trained language models in Spanish. In addition to this, our model has some cross-lingual abilities, achieving top results for English-Spanish tasks of the Linguistic Code-Switching Evaluation benchmark (LinCE) and also competitive performance against monolingual models in English Twitter tasks. To facilitate further research, we make RoBERTuito publicly available at the HuggingFace model hub together with the dataset used to pre-train it.",
99
+ }
100
+ ```