Active Learning with Multiple Views

TL;DR

This paper introduces Co-Testing, a novel multi-view active learning approach that leverages multiple and weak views to improve label efficiency across various real-world domains.

Contribution

It presents the first multi-view active learning method, Co-Testing, and extends the framework to utilize weak views for better learning performance.

Findings

Co-Testing outperforms existing active learning methods.

Effective in domains like web page classification and discourse parsing.

Utilizes multiple and weak views to enhance learning efficiency.

Abstract

Active learners alleviate the burden of labeling large amounts of data by detecting and asking the user to label only the most informative examples in the domain. We focus here on active learning for multi-view domains, in which there are several disjoint subsets of features (views), each of which is sufficient to learn the target concept. In this paper we make several contributions. First, we introduce Co-Testing, which is the first approach to multi-view active learning. Second, we extend the multi-view learning framework by also exploiting weak views, which are adequate only for learning a concept that is more general/specific than the target concept. Finally, we empirically show that Co-Testing outperforms existing active learners on a variety of real world domains such as wrapper induction, Web page classification, advertisement removal, and discourse tree parsing.