Parsley
Parsley is a unsupervised semantic classifier. Why group a bunch of questions by hand when a computer can?
Current Research and Roadmap
Currently I've implemented a simple unsupervised clustering method based off already trained word embedding models from Facebook and Google.
Cluster Your Own text
Start off by cleaning the data. This currently only supports Ahrefs keyword data.
test
parsley clean --help
Then cluster the text. You can choose which model you want to use.
parsley cluster --help
Finally, you can create a simple plot.
parsley plot --help
Sentence (or Keyword) Embeddings
- Facebook InferSent
- The Paper
- Creative Commons Non Commercial.
- Google Sentence Encoder
- The Paper
- Apache 2.0 (Mostly free to use)
Agglomeritiver Clustering
Helpful Links
- this article
- Helpfull Article
- Not a well-known author