Skip to content

Pydata 2017 workshop: build a clickbait detector with python

Notifications You must be signed in to change notification settings

cytora/clickbait-workshop

Repository files navigation

PyData 2017 Workshop

Materials for PyData workshop demonstrating the basic components of a text learning pipeline. The aim of the workshop is to build all

Installation

The easiest way to get python and the necessary packages is with conda. Go here for installation instructions. Be sure to get the python 3 versions.

Now we create a new conda environment from the environment file that contains the dependencies

$ conda env create -f environment.yml
$ source activate textml
$ python tests/test.py

Getting started

Launch the jupyter notebook to get started!

$ jupyter notebook tutorial.ipynb

In the notebook you'll learn to construct a text classification pipeline. After that you can look at the webservice directory for an example of providing machine learning as a service. The classifier is also used in advanced.ipynb to explore news data.