mnmnmnmn commited on
Commit
0189279
·
verified ·
1 Parent(s): e85d277

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +58 -0
README.md ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Open Science Dataset
2
+ ## Overview
3
+ Open Science is a large-scale, permissively licensed text dataset derived from OpenAlex, containing over 100B (105,390,332,599) words
4
+ words. OpenAlex is an open database of scholarly publications, authors, institutions, and research outputs that serves as a comprehensive source for academic literature.
5
+ ## Key Features
6
+ - **Truly Open**: Contains only permissively licensed data suitable for both commercial and non-commercial use
7
+ - **Multilingual Coverage**: Predominantly English with significant representation of [additional languages]
8
+ - **Academic Diversity**: Encompasses [types of content from OpenAlex]
9
+ ## Dataset Statistics
10
+ [Insert key statistics about:]
11
+
12
+ cc-by: 10,737,025.0 (93.09%)
13
+ 2025-01-22 16:38:32,175 - INFO - cc-by-sa: 424,556.0 (3.68%)
14
+ 2025-01-22 16:38:32,175 - INFO - public-domain: 372,583.0 (3.23%)
15
+
16
+ - Total number of documents
17
+ - Word count distribution
18
+ - Language distribution
19
+ - License type
20
+ - Temporal coverage
21
+ - [Other relevant metrics]
22
+ ## Dataset Structure
23
+ ### Data Fields
24
+ | Field | Type | Description |
25
+ |-------|------|-------------|
26
+ | [field1] | [type] | [description] |
27
+ | [field2] | [type] | [description] |
28
+ | ... | ... | ... |
29
+ ## How to Use
30
+ ### Installation
31
+ ```python
32
+ from datasets import load_dataset
33
+ dataset = load_dataset('PleIAs/openscource')
34
+ ```
35
+ ## Data Filtering Capabilities
36
+ - **Language**: Filter by specific languages using metadata
37
+ - **Temporal**: Select data based on publication dates
38
+ ## Considerations and Limitations
39
+ - While permissively licensed, users should verify compatibility with their specific use case
40
+ - Language distribution may not be uniform
41
+ - Historical texts may contain dated terminology or perspectives
42
+ - [Other important considerations]
43
+ ## Acknowledgements
44
+ This dataset was developed with support from:
45
+ (Update this, this is from common corpus)
46
+ - AI Alliance
47
+ - Jean Zay (Eviden, Idris)
48
+ - Nvidia Inception program
49
+ - Nebius AI
50
+ - Tracto AI
51
+ - Mozilla
52
+ Special recognition to:
53
+ - LANGU:IA (French Ministry of Culture and DINUM state startup)
54
+ - Alliance for Language technologies EDIC (ALT-EDIC)
55
+ - Wikimedia Enterprise
56
+ - Open science LLM community (Eleuther AI, Allen AI, HuggingFace)
57
+
58
+