EmoMeta: A Multimodal Dataset for Fine-grained Emotion Classification in Chinese Metaphors

TL;DR

EmoMeta tackles the scarcity of fine-grained, multilingual multimodal emotion datasets by constructing a Chinese dataset of 5,000 metaphorical advertisements with text-image pairs annotated for metaphor occurrence, source/target domains, and nine emotion categories plus neutral. It details a multi-source data collection pipeline, a formal annotation model, explicit emotion taxonomy, and a rigorous quality-control protocol with moderate-to-high inter-annotator agreement. The dataset reveals that fear and anticipation are prevalent in metaphorical ads, with distinct patterns for public-service and commercial content, and is publicly released to support Chinese-language metaphor and emotion research. Overall, EmoMeta provides a valuable, scalable resource to advance fine-grained emotion understanding in multimodal metaphors across languages and applications.

Abstract

Metaphors play a pivotal role in expressing emotions, making them crucial for emotional intelligence. The advent of multimodal data and widespread communication has led to a proliferation of multimodal metaphors, amplifying the complexity of emotion classification compared to single-mode scenarios. However, the scarcity of research on constructing multimodal metaphorical fine-grained emotion datasets hampers progress in this domain. Moreover, existing studies predominantly focus on English, overlooking potential variations in emotional nuances across languages. To address these gaps, we introduce a multimodal dataset in Chinese comprising 5,000 text-image pairs of metaphorical advertisements. Each entry is meticulously annotated for metaphor occurrence, domain relations and fine-grained emotion classification encompassing joy, love, trust, fear, sadness, disgust, anger, surprise, anticipation, and neutral. Our dataset is publicly accessible (https://github.com/DUTIR-YSQ/EmoMeta), facilitating further advancements in this burgeoning field.

Figures (6)

• Figure 2: An example of a metaphorical annotation
• Figure 3: Emotional distribution of EmoMeta
• Figure : (a) World No Tobacco Day, ban tobacco advertising, ensure smoke-free youth.
• Figure : (a) World No Tobacco Day, ban tobacco advertising, ensure smoke-free youth.
• Figure : (b) Low carbon emissions reduction and green living.