Alaamer commited on
Commit
30fefdd
·
verified ·
1 Parent(s): fd0af4e

Upload dataset-card.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. dataset-card.md +52 -0
dataset-card.md ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Medium Articles Dataset
2
+
3
+ ## Dataset Description
4
+
5
+ ### Dataset Summary
6
+
7
+ This dataset is a comprehensive collection of Medium articles, combining and normalizing data from multiple sources on both Kaggle and Hugging Face. The dataset ensures uniqueness of articles by removing duplicates based on article text content.
8
+
9
+ ### Languages
10
+
11
+ The dataset primarily contains articles in English.
12
+
13
+ ### Dataset Structure
14
+
15
+ The dataset is provided in Parquet format with unique entries in the text column.
16
+
17
+ ### Data Fields
18
+
19
+ - `text`: The main content of the article
20
+ - Additional fields may vary based on source datasets
21
+
22
+ ### Dataset Creation
23
+
24
+ This dataset was created by combining and normalizing multiple existing datasets from Kaggle and Hugging Face. The process includes:
25
+ 1. Downloading source datasets
26
+ 2. Normalizing data format
27
+ 3. Removing duplicate articles
28
+ 4. Handling missing values
29
+ 5. Converting to Parquet format
30
+
31
+ ### Source Data
32
+
33
+ #### Kaggle Sources:
34
+ - aiswaryaramachandran/medium-articles-with-content
35
+ - hsankesara/medium-articles
36
+ - meruvulikith/1300-towards-datascience-medium-articles-dataset
37
+
38
+ #### Hugging Face Sources:
39
+ - fabiochiu/medium-articles
40
+ - Falah/medium_articles_posts
41
+
42
+ ### Licensing Information
43
+
44
+ This dataset is released under MIT License.
45
+
46
+ ### Citation Information
47
+
48
+ If you use this dataset in your research, please cite it using the citation information in the CITATION.cff file.
49
+
50
+ ### Contributions
51
+
52
+ Thanks to all the original dataset creators and contributors. Contributions are welcome via pull requests.