Emotional Acceptance Measure (EAM): An Objective Evaluation Method Towards Information Communication Effect

Abstract

In this paper, we propose Emotional Acceptance Measure (EAM), an objective information communication effect evaluation (OICEE) method based on the theory of information entropy. Existing evaluation methods of information communication mostly utilize questionnaires and expert scoring, which often consume a lot of human resources. Aiming at this issue, we address the unexplored task - OICEE, and take the first step toward objective evaluation for assessing the effective results produced by the communication behavior in emotional dimension. Specifically, we construct a dataset for evaluating the information communication effect, design a CNN-BiGRU model based on the self-attention mechanism to calculate emotional information entropy, and propose a formula for calculating EAM score. For the first time, we introduce a novel way for objective evaluation of the information communication effect from the emotional dimension. The comparison experiment with manually annotated real evaluations shows that the EAM score achieves 94.41% correlation with subjective user evaluations, which proves the reasonableness and validity of our proposed objective evaluation method.