This repository can be used to perform active stream mining. In particular we demonstrate its use for opinion mining using the Yelp and Amazon datasets.
osm - core code
data - the downloaded datasets need to be placed in this directory.
snippets - snippets specific to preprocess the Yelp and Amazon datasets. It also includes a snippet to test the influence of oracle availability in acticve learning on the performance of the learner.
requirements.txt - project requirements
The datasets can be downloaded from the below links:
Yelp Dataset: https://www.yelp.com/dataset/challenge
Amazon Dataset: http://jmcauley.ucsd.edu/data/amazon/
More information on running the code can be obtained from the readme of the snippets