A flasky app to categorize unlabeled datasets.
In machine learning world, sometimes you need to categorize an unlabeled dataset. Probably you have obtained data from a third party and in most cases the dataset is big and you try to categorize it using an unsupervised learning algorithm such as Latent Dirichlet Allocation, KMeans, ... However, depending on the nature of your data, get a good categorization with these algorithms is really hard and you need that a human supervise the categorization. You can't infer a good classifier without a good labeled dataset.
With this purpose we have developed data-healer. A simple web application that helps you to label your datasets with a fast and friendly interface. You just have to define some configs and give it a CSV input and start to categorize. Each new categorization is registered in a new CSV with the same shape as the input one and an extra category column.
Finally, if you have tried first to infer classes with an unsupervised learning algorithm you can specify this inferred column as a default option to speed up the category selection process.
Until it will be on Pypi (Python 2 and 3 supported)::
git clone https://github.com/txerpa/data-healer
cd data-healer/
pip install .
Run data-healer::
data-healer-run
