A Dialogue Annotation Scheme for Weight Management Chat using the Trans-Theoretical Model of Health Behavior Change

TL;DR

This paper introduces a new annotation scheme for weight management dialogues based on the trans-theoretical model, and develops classifiers that improve accuracy with sentence segmentation.

Contribution

It presents a novel annotation scheme inspired by health behavior change theory and demonstrates improved classification accuracy using sentence segmentation.

Findings

Classification accuracy improves with oracle sentence segmentation

Collected and annotated human-human weight management dialogues

Proposed a trans-theoretical model-based annotation scheme

Abstract

In this study we collect and annotate human-human role-play dialogues in the domain of weight management. There are two roles in the conversation: the "seeker" who is looking for ways to lose weight and the "helper" who provides suggestions to help the "seeker" in their weight loss journey. The chat dialogues collected are then annotated with a novel annotation scheme inspired by a popular health behavior change theory called "trans-theoretical model of health behavior change". We also build classifiers to automatically predict the annotation labels used in our corpus. We find that classification accuracy improves when oracle segmentations of the interlocutors' sentences are provided compared to directly classifying unsegmented sentences.