SAD: A Stress Annotated Dataset for Recognizing Everyday Stressors in SMS-like Conversational Systems

Abstract

There is limited infrastructure for providing stress management services to those in need. To address this problem, chatbots are viewed as a scalable solution. However, one limiting factor is having clear definitions and examples of daily stress on which to build models and methods for routing appropriate advice during conversations. We developed a dataset of 6850 SMS-like sentences that can be used to classify input using a scheme of 9 stressor categories derived from: stress management literature, live conversations from a prototype chatbot system, crowdsourcing, and targeted web scraping from an online repository. In addition to releasing this dataset, we show results that are promising for classification purposes. Our contributions include: (i) a categorization of daily stressors, (ii) a dataset of SMS-like sentences, (iii) an analysis of this dataset that demonstrates its potential efficacy, and (iv) a demonstration of its utility for implementation via a simulation of model response times.