The paper describes the SemEval-2022 Task 5: Multimedia Automatic Misogyny Identification (MAMI),which explores the detection of misogynous memes on the web by taking advantage of available texts and images. The task has been organised in two related sub-tasks: the first one is focused on recognising whether a meme is misogynous or not (Sub-task A), while the second one is devoted to recognising types of misogyny (Sub-task B). MAMI has been one of the most popular tasks at SemEval-2022 with more than 400 participants, 65 teams involved in Sub-task A and 41 in Sub-task B from 13 countries. The MAMI challenge received 4214 submitted runs (of which 166 uploaded on the leader-board), denoting an enthusiastic participation for the proposed problem. The collection and annotation is described for the task dataset. The paper provides an overview of the systems proposed for the challenge, reports the results achieved in both sub-tasks and outlines a description of the main errors for a comprehension of the systems capabilities and for detailing future research perspectives.
The concept of this article is that the symbolic relationships between human beings and animals serve as a model for the relationships between the majority and the ethnic minority. We postulate that there are two representations that serve to organize these relationships between human beings and animals: a domestic and a wild one. If the domestic animal is an index of human culture, the wild animal is an index of nature which man considers himself to share with the animal. With the wild representation, contact with the animal will be taboo, as it constitutes a threat to the anthropological difference. We offer the hypothesis that ontologization of the minority, that is, the substitution of a human category with an animal category, and thus its exclusion from the human species, is a method the majority use when the taboo against contact with the wild nature is necessary. Three experiments confirm the hypothesis that the Gypsy minority (as compared with the Gadje majority) is more ontologized when the context (a monkey or a clothed dog) threatens the anthropological differentiation of the Gadje participants.
Los estudios sobre el racismo indican que en la actualidad en el contexto occidental predomina la discriminación latente sobre la manifiesta. Aún replicándose este patrón, también se observa que unas minorías son más discriminadas que otras. Se plantea que el enfoque de las representaciones sociales puede resultar más adecuado que los enfoques actitudinales para comprender esa fisonomía del racismo. Tras diferenciar discriminación y ontologización, la principal hipótesis es que la dimensión natura-cultura sirve de base para una clasificación social, dentro de la cual determinadas minorías son representadas fuera del mapa social (ontologización). Se presenta un estudio que ilustra, por una parte, cómo la cultura define la identidad humana y la natura define la identidad del animal. Y, por otra parte, que las minorías étnicas son evocadas vía la natura, positiva, del animal, como puente entre el ser humano y el animal. Se discute cómo la ontologización puede suponer una exclusión (o una no-inclusión) social sin pasar por la discriminación negativa, o sea, en el plano de la representación social, donde se recrea la exclusión con el contexto que la permite.
Stereotype is a type of social bias massively present in texts that computational models use. There are stereotypes that present special difficulties because they do not rely on personal attributes. This is the case of stereotypes about immigrants, a social category that is a preferred target of hate speech and discrimination. We propose a new approach to detect stereotypes about immigrants in texts focusing not on the personal attributes assigned to the minority but in the frames, that is, the narrative scenarios, in which the group is placed in public speeches. We have proposed a fine-grained social psychology grounded taxonomy with six categories to capture the different dimensions of the stereotype (positive vs. negative) and annotated a novel StereoImmigrants dataset with sentences that Spanish politicians have stated in the Congress of Deputies. We aggregate these categories in two supracategories: one is Victims that expresses the positive stereotypes about immigrants and the other is Threat that expresses the negative stereotype. We carried out two preliminary experiments: first, to evaluate the automatic detection of stereotypes; and second, to distinguish between the two supracategories of immigrants’ stereotypes. In these experiments, we employed state-of-the-art transformer models (monolingual and multilingual) and four classical machine learning classifiers. We achieve above 0.83 of accuracy with the BETO model in both experiments, showing that transformers can capture stereotypes about immigrants with a high level of accuracy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.