"How Was Your Weekend?" A Generative Model of Phatic Conversation

TL;DR

This paper introduces a computational model of social conversation that adheres to unspoken social norms, and demonstrates its application in a social skills training tool called EST, evaluated through a pilot study.

Contribution

The paper presents a novel generative model of social conversation that can follow or break social rules, and applies it to develop a social skills training system with initial positive results.

Findings

Model produces conversations following social norms

EST generates human-readable dialogues with guidelines

Pilot study shows model's effectiveness in social norm adherence

Abstract

Unspoken social rules, such as those that govern choosing a proper discussion topic and when to change discussion topics, guide conversational behaviors. We propose a computational model of conversation that can follow or break such rules, with participant agents that respond accordingly. Additionally, we demonstrate an application of the model: the Experimental Social Tutor (EST), a first step toward a social skills training tool that generates human-readable conversation and a conversational guideline at each point in the dialogue. Finally, we discuss the design and results of a pilot study evaluating the EST. Results show that our model is capable of producing conversations that follow social norms.