https://github.com/lostfictions/conceptnet-trim
trim conceptnet assertions to only english-language edges
https://github.com/lostfictions/conceptnet-trim
bot-food botally
Last synced: 1 day ago
JSON representation
trim conceptnet assertions to only english-language edges
- Host: GitHub
- URL: https://github.com/lostfictions/conceptnet-trim
- Owner: lostfictions
- Created: 2023-08-14T04:00:38.000Z (almost 3 years ago)
- Default Branch: master
- Last Pushed: 2023-08-14T07:16:48.000Z (almost 3 years ago)
- Last Synced: 2025-03-03T14:37:44.463Z (over 1 year ago)
- Topics: bot-food, botally
- Language: Rust
- Homepage:
- Size: 2.93 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# conceptnet-trim
trim [conceptnet](https://conceptnet.io/)'s ~34,000,000 multilingual assertions
(about 10gb of tsv) into a tidy ~3,400,000 english-language assertions (in json
format).
1. clone this repo
2. [download the latest version of
conceptnet](https://github.com/commonsense/conceptnet5/wiki/Downloads) (5.7.0
at the time of writing)
3. extract it to `data/assertions.csv` in the root of this repo
4. run `cargo run -r` to run in release mode. the trimmed assertions will be
written to `data/trimmed.json`.
or download a pre-trimmed file from the [releases
page](https://github.com/lostfictions/conceptnet-trim/releases).