Datasets:
Electrotubbie
commited on
Commit
•
044f4fc
1
Parent(s):
4591bd8
edit readme
Browse files
README.md
CHANGED
@@ -7,4 +7,32 @@ language:
|
|
7 |
- ky
|
8 |
size_categories:
|
9 |
- 100K<n<1M
|
10 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
- ky
|
8 |
size_categories:
|
9 |
- 100K<n<1M
|
10 |
+
---
|
11 |
+
|
12 |
+
## Description
|
13 |
+
A dataset with texts and the categories to which these texts belong.
|
14 |
+
|
15 |
+
## Usage
|
16 |
+
This dataset can be used to check language models for the correct classification of texts by category.
|
17 |
+
|
18 |
+
## Dataset structure:
|
19 |
+
**lang**: the language to which the text source belongs;
|
20 |
+
**title**: the title of the text;
|
21 |
+
**original_text**: original text taken from a web page;
|
22 |
+
**processed_text**: processed text using preprocessing functions;
|
23 |
+
**category**: the category to which the text belongs;
|
24 |
+
**processed**: flag indicating that one or more sentence has been deleted from the text;
|
25 |
+
**url**: link to the source;
|
26 |
+
**date**: date of publication of the text;
|
27 |
+
|
28 |
+
## The creation process
|
29 |
+
This dataset was obtained by parsing news resources of countries and regions of native speakers of Turkic languages, such as Bashkir, Kazakh and Kyrgyz.
|
30 |
+
During parsing, it was a priori believed that the language of the articles was written in the language of the region about which the news was written.
|
31 |
+
After parsing, the text of the articles was processed through the preprocessing functions described on [github](https://github.com/Electrotubbie/turk_langs_analyse ).
|
32 |
+
|
33 |
+
The scheme of text preprocessing and validation is as follows:
|
34 |
+
- cleaning the text from unnecessary constructions using regular expressions;
|
35 |
+
- splitting text into sentences using the sentenize function of the razdel module;
|
36 |
+
- making predictions for each sentence using the lid.176.bin model, as well as the fasttext module;
|
37 |
+
- deleting sentences written in non-Turkic languages;
|
38 |
+
- combining valid sentences into text and getting the processed_text column.
|