Hi-EF: Benchmarking Emotion Forecasting in Human-interaction

TL;DR

This paper introduces a new benchmark for emotion forecasting in human interactions, focusing on predicting emotional responses based on partner cues to improve accuracy and understanding.

Contribution

It defines the task of Human-interaction-based Emotion Forecasting (EF) and provides a structured framework for analyzing emotional dynamics in two-party interactions.

Findings

Defines the EF task within human interactions

Proposes a benchmark dataset for emotion forecasting

Highlights the influence of partner cues on emotional predictions

Abstract

Affective Forecasting is an psychology task that involves predicting an individual's future emotional responses, often hampered by reliance on external factors leading to inaccuracies, and typically remains at a qualitative analysis stage. To address these challenges, we narrows the scope of Affective Forecasting by introducing the concept of Human-interaction-based Emotion Forecasting (EF). This task is set within the context of a two-party interaction, positing that an individual's emotions are significantly influenced by their interaction partner's emotional expressions and informational cues. This dynamic provides a structured perspective for exploring the patterns of emotional change, thereby enhancing the feasibility of emotion forecasting.