In this repository, we have implemented semi-supervised methods for text classification which can be deployed in SPARK distributed system.
Under the TwitterPreprocessing, we have implemented the text preprocessing part of our process.
Self-Learning, Co-Training classification have been implemented for textual classification.
Implementation link
Derived labels link
Crowdsourced labels link
Promo video link
Paper link
In order to cite our work please use this reference:
@inproceedings{iosifidis2017large, title={Large scale sentiment learning with limited labels}, author={Iosifidis, Vasileios and Ntoutsi, Eirini}, booktitle={Proceedings of the 23rd ACM SIGKDD international conference on knowledge discovery and data mining}, pages={1823--1832}, year={2017}, organization={ACM} }