Understanding the sources of variability in annotations is crucial for developing fair NLP systems, especially for tasks like sexism detection where demographic bias is a concern. This study investigates the extent to which annotator demographic features influence labeling decisions compared to text content. Using a Generalized Linear Mixed Model, we quantify this influence, finding that while statistically present, demographic factors account for a minor fraction (~8%) of the observed variance, with tweet content being the dominant factor. We then assess the reliability of Generative AI (GenAI) models as annotators, specifically evaluating if guiding them with demographic personas improves alignment with human judgments. Our results indicate that simplistic persona prompting often fails to enhance, and sometimes degrades, performance compared to baseline models. Furthermore, explainable AI (XAI) techniques reveal that model predictions rely heavily on content-specific tokens related to sexism, rather than correlates of demographic characteristics. We argue that focusing on content-driven explanations and robust annotation protocols offers a more reliable path towards fairness than potentially persona simulation.

Text-based personality computing (TPC) has gained many research interests in NLP. In this paper, we describe 15 challenges that we consider deserving the attention of the NLP research community. These challenges are organized by the following topics: personality taxonomies, measurement quality, datasets, performance evaluation, modelling choices, as well as ethics and fairness. When addressing each challenge, not only do we combine perspectives from both NLP and social sciences, but also offer concrete suggestions. We hope to inspire more valid and reliable TPC research.

This paper describes the system submitted by the PRHLT-UPV team for the task 8 of SemEval-2020: Memotion Analysis. We propose a multimodal model that combines pretrained models of the BERT and VGG architectures. The BERT model is used to process the textual information and VGG the images. The multimodal model is used to classify memes according to the presence of offensive, sarcastic, humorous and motivating content. Also, a sentiment analysis of memes is carried out with the proposed model. In the experiments, the model is compared with other approaches to analyze the relevance of the multimodal model. The results show encouraging performances on the final leaderboard of the competition, reaching good positions in the ranking of systems.

This paper describes the participation of USI-IR in WASSA 2018 Implicit Emotion Shared Task. We propose a relevance feedback approach employing a sequential model (biLSTM) and word embeddings derived from a large collection of tweets. To this end, we assume that the top-k predictions produce at a first classification step are correct (based on the model accuracy) and use them as new examples to re-train the network.