https://github.com/datadotworld/foia-app
R Shiny App created to predict the success rate of Freedom of Information Act requests.
https://github.com/datadotworld/foia-app
dwstruct-t90-deprecated r shiny
Last synced: 11 months ago
JSON representation
R Shiny App created to predict the success rate of Freedom of Information Act requests.
- Host: GitHub
- URL: https://github.com/datadotworld/foia-app
- Owner: datadotworld
- Archived: true
- Created: 2017-05-12T14:52:56.000Z (about 9 years ago)
- Default Branch: master
- Last Pushed: 2017-12-11T15:44:31.000Z (over 8 years ago)
- Last Synced: 2025-05-08T21:14:53.239Z (about 1 year ago)
- Topics: dwstruct-t90-deprecated, r, shiny
- Language: R
- Homepage:
- Size: 7.31 MB
- Stars: 16
- Watchers: 40
- Forks: 4
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# foia_shiny_app
This R Shiny App was created to predict the success rate of Freedom of Information Act requests based on a dataset of 9,000+ previous FOIA requests and their outcomes.
[View the data on data.world](https://data.world/rdowns26/foia-analysis)
## Methodology
This model uses K Nearest Neighbors with k=5. First, it reads in the previous reqests that have already had the important features extracted ([see feature extraction python notebook](https://data.world/rdowns26/foia-analysis/file/FOIA_Feature_Extraction.ipynb)) and builds a model. Then, it takes the user's text input and performs the same feature extraction in real time. It also translates the user's agency into a boolean variable based on whether that agency has a success rate > 50%. Then the model outputs a predicted likelihood of success, as well as the features the script extracted from the text input.
**Have ideas on how to make this model better? We'd love your help! Submit a pull request or [request to be a contributor on data.world.](https://data.world/rdowns26/foia-analysis/contributors)**