Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,75 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-4.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# ShopTC-100K Dataset
|
| 6 |
+
|
| 7 |
+

|
| 8 |
+
|
| 9 |
+
The ShopTC-100K dataset is collected using TermMiner, a data collection and topic modeling pipeline introduced in the paper:
|
| 10 |
+
|
| 11 |
+
**Harmful Terms and Where to Find Them: Measuring and Modeling Unfavorable Financial Terms and Conditions in Shopping Websites at Scale**
|
| 12 |
+
|
| 13 |
+
To cite this dataset and related research, please use the following reference:
|
| 14 |
+
```
|
| 15 |
+
@inproceedings{tsai2025harmful,
|
| 16 |
+
author = {Elisa Tsai and Neal Mangaokar and Boyuan Zheng and Haizhong Zheng and Atul Prakash},
|
| 17 |
+
title = {Harmful Terms and Where to Find Them: Measuring and Modeling Unfavorable Financial Terms and Conditions in Shopping Websites at Scale},
|
| 18 |
+
booktitle = {Proceedings of the ACM Web Conference 2025 (WWW β25)},
|
| 19 |
+
year = {2025},
|
| 20 |
+
location = {Sydney, NSW, Australia},
|
| 21 |
+
publisher = {ACM},
|
| 22 |
+
address = {New York, NY, USA},
|
| 23 |
+
pages = {14},
|
| 24 |
+
month = {April 28-May 2},
|
| 25 |
+
doi = {10.1145/3696410.3714573}
|
| 26 |
+
}
|
| 27 |
+
```
|
| 28 |
+
|
| 29 |
+
## Dataset Description
|
| 30 |
+
|
| 31 |
+
The dataset consists of sanitized terms extracted from 8,251 e-commerce websites with English-language terms and conditions. The websites were sourced from the [Tranco list](https://tranco-list.eu/) (as of April 2024). The dataset contains:
|
| 32 |
+
|
| 33 |
+
- 1,825,231 sanitized sentences
|
| 34 |
+
- 7,777 unique websites
|
| 35 |
+
- Four split files for ease of use:
|
| 36 |
+
```
|
| 37 |
+
ShopTC-100K
|
| 38 |
+
βββ sanitized_split1.csv
|
| 39 |
+
βββ sanitized_split2.csv
|
| 40 |
+
βββ sanitized_split3.csv
|
| 41 |
+
βββ sanitized_split4.csv
|
| 42 |
+
```
|
| 43 |
+
|
| 44 |
+
### Data Sanitization Process
|
| 45 |
+
|
| 46 |
+
The extracted terms are cleaned and structured using a multi-step sanitization pipeline:
|
| 47 |
+
|
| 48 |
+
- HTML Parsing: Raw HTML content is processed to extract text from `<p>` tags.
|
| 49 |
+
- Sentence Tokenization: Text is split into sentences using a transformer-based tokenization model.
|
| 50 |
+
- Filtering: Short sentences (<10 words) and duplicates are removed.
|
| 51 |
+
- Preprocessing: Newline characters and extra whitespace are cleaned.
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
| Split File | Rows | Columns | Unique Websites |
|
| 55 |
+
|--------------------------------------|---------|---------|----------------|
|
| 56 |
+
| sanitized_split1.csv | 523,760 | 2 | 1,979 |
|
| 57 |
+
| sanitized_split2.csv | 454,966 | 2 | 1,973 |
|
| 58 |
+
| sanitized_split3.csv | 425,028 | 2 | 1,988 |
|
| 59 |
+
| sanitized_split4.csv | 421,477 | 2 | 1,837 |
|
| 60 |
+
|
| 61 |
+
### Example Data
|
| 62 |
+
|
| 63 |
+
The dataset is structured as follows:
|
| 64 |
+
| URL | Paragraph |
|
| 65 |
+
|-----------------------|----------------------------------------------------------------|
|
| 66 |
+
| pythonanywhere.com | Copyright Β© 2011-2024 PythonAnywhere LLP β Terms of Service apply. |
|
| 67 |
+
| pythonanywhere.com | We use cookies to provide social media features and to analyze our traffic. |
|
| 68 |
+
| pythonanywhere.com | 2.8 You acknowledge that clicking on Links may lead to third-party sites. |
|
| 69 |
+
| pythonanywhere.com | 3.4 No payment will be made unless and until Account verification is complete. |
|
| 70 |
+
| pythonanywhere.com | 11.3 All licenses granted to you in this agreement are non-transferable. |
|
| 71 |
+
|
| 72 |
+
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
|