More Is Better: Large Scale Partially-supervised Sentiment Classification - Appendix

TL;DR

This paper presents a bootstrapping algorithm for sentiment classification that leverages large-scale partially labeled data, significantly improving performance in various semi-supervised and domain adaptation tasks.

Contribution

It introduces a novel bootstrapping method for learning from partially labeled data and demonstrates its effectiveness on large-scale Amazon review datasets.

Findings

Reduced test error by over 50% in some cases

Effective semi-supervised learning with 15 million reviews

Improved domain adaptation performance

Abstract

We describe a bootstrapping algorithm to learn from partially labeled data, and the results of an empirical study for using it to improve performance of sentiment classification using up to 15 million unlabeled Amazon product reviews. Our experiments cover semi-supervised learning, domain adaptation and weakly supervised learning. In some cases our methods were able to reduce test error by more than half using such large amount of data. NOTICE: This is only the supplementary material.