https://github.com/renyuneyun/pp-analyzer
https://github.com/renyuneyun/pp-analyzer
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/renyuneyun/pp-analyzer
- Owner: renyuneyun
- License: apache-2.0
- Created: 2024-10-01T19:27:16.000Z (almost 2 years ago)
- Default Branch: master
- Last Pushed: 2025-05-15T20:54:43.000Z (about 1 year ago)
- Last Synced: 2025-07-14T01:34:30.686Z (about 1 year ago)
- Language: Jupyter Notebook
- Size: 1010 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
README
## PoliAnalyzer
This is the repo for PoliAnalyzer, a semantic-centric tool to automatically analyze privacy policies. Conceptually, it contains the following components:
- NLP workflow: Convert from natural-language privacy policies to KG ($Pr^2Graph$), grounded in DPV
- LLM pipeline: The key driver for the NLP workflow
- Formal policy converter: Convert $Pr^2Graph$ to formal policies, namely [ODRL](https://www.w3.org/TR/odrl-model/) and [DToU](https://github.com/OxfordHCC/solid-dtou).
- This is performed through SPARQL queries. See `pp-analyze/README.md` for further information
- User profile analyzer: Check the compliance of users' preferences against formal policies
> The LLM pipeline is in `fine-tune/` dir; the rest is in `pp-analyze/` dir.
Please refer to individual sub-directories for detailed instructions on how to use them.
At the moment, there is no GUI. Jupyter notebooks are provided with relevant documentation to explain how to use the system.
## Content
- `fine-tune`: Tasks for fine-tuning models and evaluating the LLM pipeline are contained in this directory.
- `pp-analyze`: This directory contains user-facing or artifact-creating functionalities of PoliAnalyzer.
- `ppa-commons`: Shared code between `fine-tune` and `pp-analyzer`.
- `visualize`: Code for generating figures for visualization. Mainly for development use, and not generally useful for users.
## Additional resources
There are some additional resources that may be useful for users browsing this repo, such as by pulling in necessary benchmark datasets or formal policies for compatibility checking.
- Enriched Policy-IE: \url{https://doi.org/10.5281/zenodo.15392162}
- $Pr^2Graph$ and formal policies in ODRL and DToU for top-100 websites: \url{https://doi.org/10.5281/zenodo.20081062}
## License
Unless otherwise stated, all source codes are licensed under the Apache-2.0 license.
## Citation guideline
The other resources have their own respective citation guidelines.
To cite this repository, use the following snippet:
```
@misc{zhao_renyuneyun_2026,
title = {Renyuneyun/Pp-Analyzer},
author = {Zhao, Rui},
year = 2026,
month = may,
url = {https://github.com/renyuneyun/pp-analyzer},
urldate = {2026-05-08},
copyright = {Apache-2.0},
keywords = {/unread}
}
```
or, if not using BibTeX, you can use the following:
```
Zhao, Rui. Renyuneyun/Pp-Analyzer. October 1, 2024, https://github.com/renyuneyun/pp-analyzer.
```