To effectively address global environmental challenges, we must have tools that allow us to carefully monitor how citizens, policy makers and other stakeholders debate sustainability. However, there are currently very few NLP resources and tools specialized for this topic. This paper presents EnviS, a multilingual corpus (Italian, English, and Indonesian) for investigating the debate on environmental sustainability in social media using Structured Sentiment Analysis. We introduce a framework for the automatic aggregation of span-level annotations that preserves the annotators’ perspective and avoids manual intervention by safeguarding the quality of the annotations. We performed a series of experiments with four open-source instruction-based Large Language Models in zero-shot and few-shot settings, where we have measures the impact of the order and number of shots. The results further confirm the ineffectiveness of LLMs in extracting fine-grained sentiment information, being outperformed by a supervised state-of-the-art neural method trained on very few data. This questions the suitability of LLMs for rich knowledge/information extraction tasks requiring manipulation of text spans. In particular, our error analysis indicates that LLMs mostly struggle in identifying the sentiment term or its associated polarity, failing to extract full sentiment triples.

We present EPIC (English Perspectivist Irony Corpus), the first annotated corpus for irony analysis based on the principles of data perspectivism. The corpus contains short conversations from social media in five regional varieties of English, and it is annotated by contributors from five countries corresponding to those varieties. We analyse the resource along the perspectives induced by the diversity of the annotators, in terms of origin, age, and gender, and the relationship between these dimensions, irony, and the topics of conversation. We validate EPIC by creating perspective-aware models that encode the perspectives of annotators grouped according to their demographic characteristics. Firstly, the performance of perspectivist models confirms that different annotators induce very different models. Secondly, in the classification of ironic and non-ironic texts, perspectivist models prove to be generally more confident than the non-perspectivist ones. Furthermore, comparing the performance on a perspective-based test set with those achieved on a gold standard test set, we can observe how perspectivist models tend to detect more precisely the positive class, showing their ability to capture the different perceptions of irony. Thanks to these models, we are moreover able to show interesting insights about the variation in the perception of irony by the different groups of annotators, such as among different generations and nationalities.

In this paper we present a set of multilingual experiments tackling the task of Stance Detection in five different languages: English, Spanish, Catalan, French and Italian. Furthermore, we study the phenomenon of stance with respect to six different targets – one per language, and two different for Italian – employing a variety of machine learning algorithms that primarily exploit morphological and syntactic knowledge as features, represented throughout the format of Universal Dependencies. Results seem to suggest that the methodology employed is not beneficial per se, but might be useful to exploit the same features with a different methodology.