Commit ·
82bc507
1
Parent(s): 24fffe4
Update README.md
Browse files
README.md
CHANGED
|
@@ -39,6 +39,10 @@ This dataset is a combination of data from various reputable sources, each contr
|
|
| 39 |
## Usage
|
| 40 |
This dataset can be used for pre-training language models on financial tweets.
|
| 41 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
## Acknowledgements
|
| 43 |
|
| 44 |
We extend our heartfelt gratitude to all the authors and contributors of the original datasets. Their efforts in data collection and curation have been pivotal in creating this comprehensive resource.
|
|
|
|
| 39 |
## Usage
|
| 40 |
This dataset can be used for pre-training language models on financial tweets.
|
| 41 |
|
| 42 |
+
## Pre-processing Steps
|
| 43 |
+
Originally the combined datasets consist of 9,336,675 rows. However, there are some duplicates and not useful tweets in there. The dataset has been cleaned of `t.co` URLs, duplicate text, empty text, and tweets that end with '...'.
|
| 44 |
+
As a result the cleaned dataset consist of 8,024,269 rows, which is the one available here.
|
| 45 |
+
|
| 46 |
## Acknowledgements
|
| 47 |
|
| 48 |
We extend our heartfelt gratitude to all the authors and contributors of the original datasets. Their efforts in data collection and curation have been pivotal in creating this comprehensive resource.
|