# On the Complementarity of Images and Text for the Expression of Emotions in Social Media

arXiv: https://arxiv.org/abs/2202.07427 · HTML: https://mlsanity.com/paper/2202.07427/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 58 references that are in the map.
- First posted on arXiv 11 Feb 2022.

## Builds on (most cited first, 17 of 17)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [The Pushshift Reddit Dataset](https://mlsanity.com/paper/2001.08435/) (2020)
- [Deep Facial Expression Recognition: A Survey](https://mlsanity.com/paper/1804.08348/) (2018)
- [Emotion-Cause Pair Extraction: A New Task to Emotion Analysis in Texts](https://mlsanity.com/paper/1906.01267/) (2019)
- [Integrating Text and Image: Determining Multimodal Document Intent in Instagram Posts](https://mlsanity.com/paper/1904.09073/) (2019)
- [GoodNewsEveryone: A Corpus of News Headlines Annotated with Emotions, Semantic Roles, and Reader Perception](https://mlsanity.com/paper/1912.03184/) (2019)
- [Appraisal Theories for Emotion Classification in Text](https://mlsanity.com/paper/2003.14155/) (2020)
- [A Question Answering Approach to Emotion Cause Extraction](https://mlsanity.com/paper/1708.05482/) (2017)
- [Equal But Not The Same: Understanding the Implicit Relationship Between Persuasive Images and Text](https://mlsanity.com/paper/1807.08205/) (2018)
- [Emotion Recognition under Consideration of the Emotion Component Process Model](https://mlsanity.com/paper/2107.12895/) (2021)
- [IEST: WASSA-2018 Implicit Emotions Shared Task](https://mlsanity.com/paper/1809.01083/) (2018)
- [Token Sequence Labeling vs. Clause Classification for English Emotion Stimulus Detection](https://mlsanity.com/paper/2010.07557/) (2020)
- [Detecting Concept-level Emotion Cause in Microblogging](https://mlsanity.com/paper/1504.08050/) (2015)
- [When Saliency Meets Sentiment: Understanding How Image Content Invokes Emotion and Sentiment](https://mlsanity.com/paper/1611.04636/) (2016)
- [Knowledge-rich Image Gist Understanding Beyond Literal Meaning](https://mlsanity.com/paper/1904.08709/) (2019)

## Built on by (most cited first, 1 of 1)

- [Where are We in Event-centric Emotion Analysis? Bridging Emotion Role Labeling and Appraisal-based Approaches](https://mlsanity.com/paper/2309.02092/) (2023)

## Abstract (arXiv)

Authors of posts in social media communicate their emotions and what causes them with text and images. While there is work on emotion and stimulus detection for each modality separately, it is yet unknown if the modalities contain complementary emotion information in social media. We aim at filling this research gap and contribute a novel, annotated corpus of English multimodal Reddit posts. On this resource, we develop models to automatically detect the relation between image and text, an emotion stimulus category and the emotion class. We evaluate if these tasks require both modalities and find for the image-text relations, that text alone is sufficient for most categories (complementary, illustrative, opposing): the information in the text allows to predict if an image is required for emotion understanding. The emotions of anger and sadness are best predicted with a multimodal model, while text alone is sufficient for disgust, joy, and surprise. Stimuli depicted by objects, animals, food, or a person are best predicted by image-only models, while multimodal models are most effective on art, events, memes, places, or screenshots.