Professor in Political Science and Computer and Information Science

David Lazer

ConStance: Modeling Annotation Contexts to Improve Stance Classification

Peer Reviewed Computer Science Conference
Publication date: 
09/2017
Authors: 
Kenneth Joseph
Lisa Friedland
Will Hobbs
David Lazer
Oren Tsur
ConStance: Modeling Annotation Contexts to Improve Stance Classification

Manual annotations are a prerequisite for many applications of machine learning. However, weaknesses in the annotation process itself are easy to overlook. In particular, scholars often choose what information to give to annotators without examining these decisions empirically. For subjective tasks such as sentiment analysis, sarcasm, and stance detection, such choices can impact results. Here, for the task of political stance detection on Twitter, we show that providing too little context can result in noisy and uncertain annotations, whereas providing too strong a context may cause it to outweigh other signals. To characterize and reduce these biases, we develop ConStance, a general model for reasoning about annotations across information conditions. Given conflicting labels produced by multiple annotators seeing the same instances with different contexts, ConStance simultaneously estimates gold standard labels and also learns a classifier for new instances. We show that the classifier learned by ConStance outperforms a variety of baselines at predicting political stance, while the model's interpretable parameters shed light on the effects of each context.

Research Areas TOC

Computational Social Science, 21st Century Democracy, Political Networks

Computational Social Science, Collective Cognition

DNA and the Criminal Justice System

21st Century Democracy, Political Networks