Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,29 @@
|
|
1 |
---
|
2 |
license: cc
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: cc
|
3 |
+
task_categories:
|
4 |
+
- question-answering
|
5 |
+
language:
|
6 |
+
- en
|
7 |
---
|
8 |
+
|
9 |
+
# Dataset Card for WikiDoc
|
10 |
+
For the dataset containing rephrased content from the living textbook refer to [this dataset](https://huggingface.co/datasets/medalpaca/medical_meadow_wikidoc)
|
11 |
+
## Dataset Description
|
12 |
+
|
13 |
+
- **Source:** https://www.wikidoc.org/index.php/Main_Page
|
14 |
+
- **Repository:** https://github.com/kbressem/medalpaca
|
15 |
+
- **Paper:** TBA
|
16 |
+
|
17 |
+
### Dataset Summary
|
18 |
+
This dataset containes medical question-answer pairs extracted from [WikiDoc](https://www.wikidoc.org/index.php/Main_Page),
|
19 |
+
a collaborative platform for medical professionals to share and contribute to up-to-date medical knowledge.
|
20 |
+
The platform has to main subsites, the "Living Textbook" and "Patient Information". The "Living Textbook"
|
21 |
+
contains chapters for various medical specialties, which we crawled. We then used GTP-3.5-Turbo to rephrase
|
22 |
+
the paragraph heading to a question and used the paragraph as answer. Patient Information is structured differently,
|
23 |
+
in that each section subheading is already a question, making rephrasing them obsolete.
|
24 |
+
|
25 |
+
**Note:** This dataset is still a WIP. While the Q/A pairs from the patient information seems to be mostly correct,
|
26 |
+
the conversion using GPT-3.5-Turbo yielded some unsatisfactory results in approximately 30% of cases. We are in the process of cleaning this dataset.
|
27 |
+
|
28 |
+
### Citation Information
|
29 |
+
TBA
|