People worldwide use language in subtle and complex ways to express emotions. Although emotion recognition–an umbrella term for several NLP tasks–impacts various applications within NLP and beyond, most work in this area has focused on high-resource languages. This has led to significant disparities in research efforts and proposed solutions, particularly for under-resourced languages, which often lack high-quality annotated datasets.In this paper, we present BRIGHTER–a collection of multi-labeled, emotion-annotated datasets in 28 different languages and across several domains. BRIGHTER primarily covers low-resource languages from Africa, Asia, Eastern Europe, and Latin America, with instances labeled by fluent speakers. We highlight the challenges related to the data collection and annotation processes, and then report experimental results for monolingual and crosslingual multi-label emotion identification, as well as emotion intensity recognition. We analyse the variability in performance across languages and text domains, both with and without the use of LLMs, and show that the BRIGHTER datasets represent a meaningful step towards addressing the gap in text-based emotion recognition.

Large Language Models (LLMs) excel in zero-shot and few-shot tasks, but achieving similar performance with encoder-only models like BERT and RoBERTa has been challenging due to their architecture. However, encoders offer advantages such as lower computational and memory costs. Recent work adapts them for zero-shot generalization using Statement Tuning, which reformulates tasks into finite templates. We extend this approach to multilingual NLP, exploring whether encoders can achieve zero-shot cross-lingual generalization and serve as efficient alternatives to memory-intensive LLMs for low-resource languages. Our results show that state-of-the-art encoder models generalize well across languages, rivaling multilingual LLMs while being more efficient. We also analyze multilingual Statement Tuning dataset design, efficiency gains, and language-specific generalization, contributing to more inclusive and resource-efficient NLP models. We release our code and models.

When we encountered upsetting or tragic situations involving other people, we might feel certain emotions that are congruent, though not necessarily identical, to what that person might went through. These kind of vicarious emotions are what defined empathy and distress, they can be seen as a form of emotional response to other people in need. In this paper, we describe our participation in WASSA 2024 Shared Task 3 in predicting writer’s level of empathy and distress from their personal essays. We approach this task by assuming one’s level of empathy and distress can be revealed from the emotional patterns within their essay. By extracting the emotional patterns from essays via an emotion classifier, we regress the empathy and distress levels from these patterns. Through correlation and model explainability analysis, we found that there are similar set of emotions, such as sadness or disappointment, and distinct set of emotions, such as anger or approval, that might describe the writer’s level of empathy and distress. We hope that our approach and findings could serve as a basis for future work that try to model and explain empathy and distress from emotional patterns.