Large Language Models (LLMs) reproduce and exacerbate the social biases present in their training data, and resources to quantify this issue are limited. While research has attempted to identify and mitigate such biases, most efforts have been concentrated around English, lagging the rapid advancement of LLMs in multilingual settings. In this paper, we introduce a new multilingual parallel dataset SHADES to help address this issue, designed for examining culturally-specific stereotypes that may be learned by LLMs. The dataset includes stereotypes from 20 regions around the world and 16 languages, spanning multiple identity categories subject to discrimination worldwide. We demonstrate its utility in a series of exploratory evaluations for both “base” and “instruction-tuned” language models. Our results suggest that stereotypes are consistently reflected across models and languages, with some languages and models indicating much stronger stereotype biases than others.

This paper presents MoCCA, a Model of Comparative Concepts for Aligning Constructicons under development by a consortium of research groups building Constructicons of different languages including Brazilian Portuguese, English, German and Swedish. The Constructicons will be aligned by using comparative concepts (CCs) providing language-neutral definitions of linguistic properties. The CCs are drawn from typological research on grammatical categories and constructions, and from FrameNet frames, organized in a conceptual network. Language-specific constructions are linked to the CCs in accordance with general principles. MoCCA is organized into files of two types: a largely static CC Database file and multiple Linking files containing relations between constructions in a Constructicon and the CCs. Tools are planned to facilitate visualization of the CC network and linking of constructions to the CCs. All files and guidelines will be versioned, and a mechanism is set up to report cases where a language-specific construction cannot be easily linked to existing CCs.

This paper presents the Frame2 dataset, a multimodal dataset built from a corpus of a Brazilian travel TV show annotated for FrameNet categories for both the text and image communicative modes. Frame2 comprises 230 minutes of video, which are correlated with 2,915 sentences either transcribing the audio spoken during the episodes or the subtitling segments of the show where the host conducts interviews in English. For this first release of the dataset, a total of 11,796 annotation sets for the sentences and 6,841 for the video are included. Each of the former includes a target lexical unit evoking a frame or one or more frame elements. For each video annotation, a bounding box in the image is correlated with a frame, a frame element and lexical unit evoking a frame in FrameNet.

This paper presents Framed Multi30K (FM30K), a novel frame-based Brazilian Portuguese multimodal-multilingual dataset which i) extends the Multi30K dataset (Elliot et al., 2016) with 158,915 original Brazilian Portuguese descriptions, and 30,104 Brazilian Portuguese translations from original English descriptions; and ii) adds 2,677,613 frame evocation labels to the 158,915 English descriptions and to the ones created for Brazilian Portuguese; (iii) extends the Flickr30k Entities dataset (Plummer et al., 2015) with 190,608 frames and Frame Elements correlations with the existing phrase-to-region correlations.

This paper reports on negative results in a task of automatic identification of schematic clausal constructions and their elements in Brazilian Portuguese. The experiment was set up so as to test whether form and meaning properties of constructions, modeled in terms of Universal Dependencies and FrameNet Frames in a Constructicon, would improve the performance of transformer models in the task. Qualitative analysis of the results indicate that alternatives to the linearization of those properties, dataset size and a post-processing module should be explored in the future as a means to make use of information in Constructicons for NLP tasks.

This paper presents Charon, a web tool for annotating multimodal corpora with FrameNet categories. Annotation can be made for corpora containing both static images and video sequences paired – or not – with text sequences. The pipeline features, besides the annotation interface, corpus import and pre-processing tools.

In this paper we present Scylla, a methodology for domain adaptation of Neural Machine Translation (NMT) systems that make use of a multilingual FrameNet enriched with qualia relations as an external knowledge base. Domain adaptation techniques used in NMT usually require fine-tuning and in-domain training data, which may pose difficulties for those working with lesser-resourced languages and may also lead to performance decay of the NMT system for out-of-domain sentences. Scylla does not require fine-tuning of the NMT model, avoiding the risk of model over-fitting and consequent decrease in performance for out-of-domain translations. Two versions of Scylla are presented: one using the source sentence as input, and another one using the target sentence. We evaluate Scylla in comparison to a state-of-the-art commercial NMT system in an experiment in which 50 sentences from the Sports domain are translated from Brazilian Portuguese to English. The two versions of Scylla significantly outperform the baseline commercial system in HTER.

Frame shift is a cross-linguistic phenomenon in translation which results in corresponding pairs of linguistic material evoking different frames. The ability to predict frame shifts would enable (semi-)automatic creation of multilingual frame annotations and thus speeding up FrameNet creation through annotation projection. Here, we first characterize how frame shifts result from other linguistic divergences such as translational divergences and construal differences. Our analysis also shows that many pairs of frames in frame shifts are multi-hop away from each other in Berkeley FrameNet’s net-like configuration. Then, we propose the Frame Shift Prediction task and demonstrate that our graph attention networks, combined with auxiliary training, can learn cross-linguistic frame-to-frame correspondence and predict frame shifts.

This paper presents Lutma, a collaborative, semi-constrained, tutorial-based tool for contributing frames and lexical units to the Global FrameNet initiative. The tool parameterizes the process of frame creation, avoiding consistency violations and promoting the integration of frames contributed by the community with existing frames. Lutma is structured in a wizard-like fashion so as to provide users with text and video tutorials relevant for each step in the frame creation process. We argue that this tool will allow for a sensible expansion of FrameNet coverage in terms of both languages and cultural perspectives encoded by them, positioning frames as a viable alternative for representing perspective in language models.

This paper argues in favor of the adoption of annotation practices for multimodal datasets that recognize and represent the inherently perspectivized nature of multimodal communication. To support our claim, we present a set of annotation experiments in which FrameNet annotation is applied to the Multi30k and the Flickr 30k Entities datasets. We assess the cosine similarity between the semantic representations derived from the annotation of both pictures and captions for frames. Our findings indicate that: (i) frame semantic similarity between captions of the same picture produced in different languages is sensitive to whether the caption is a translation of another caption or not, and (ii) picture annotation for semantic frames is sensitive to whether the image is annotated in presence of a caption or not.

Human speakers have an extensive toolkit of ways to express themselves. In this paper, we engage with an idea largely absent from discussions of meaning in natural language understanding—namely, that the way something is expressed reflects different ways of conceptualizing or construing the information being conveyed. We first define this phenomenon more precisely, drawing on considerable prior work in theoretical cognitive semantics and psycholinguistics. We then survey some dimensions of construed meaning and show how insights from construal could inform theoretical and practical work in NLP.

Framenets as an incarnation of frame semantics have been set up to deal with lexicographic issues (cf. Fillmore and Baker 2010, among others). They are thus concerned with lexical units (LUs) and the conceptual structure which categorizes these together. These lexically-evoked frames, however, do not reflect pragmatic properties of constructions (LUs and other types of constructions), such as expressing illocutions or being considered polite or very informal. From the viewpoint of a multilingual annotation effort, the Global FrameNet Shared Annotation Task, we discuss two phenomena, greetings and tag questions, which highlight the necessity both to investigate the role between construction and frame annotation on the one hand and to develop pragmatic frames describing social interactions which are not explicitly lexicalized.

Multimodal aspects of human communication are key in several applications of Natural Language Processing, such as Machine Translation and Natural Language Generation. Despite recent advances in integrating multimodality into Computational Linguistics, the merge between NLP and Computer Vision techniques is still timid, especially when it comes to providing fine-grained accounts for meaning construction. This paper reports on research aiming to determine appropriate methodology and develop a computational tool to annotate multimodal corpora according to a principled structured semantic representation of events, relations and entities: FrameNet. Taking a Brazilian television travel show as corpus, a pilot study was conducted to annotate the frames that are evoked by the audio and the ones that are evoked by visual elements. We also implemented a Multimodal Annotation tool which allows annotators to choose frames and locate frame elements both in the text and in the images, while keeping track of the time span in which those elements are active in each modality. Results suggest that adding a multimodal domain to the linguistic layer of annotation and analysis contributes both to enrich the kind of information that can be tagged in a corpus, and to enhance FrameNet as a model of linguistic cognition.

Although FrameNet is recognized as one of the most fine-grained lexical databases, its coverage of lexical units is still limited. To tackle this issue, we propose a two-step frame induction process: for a set of lexical units not yet present in Berkeley FrameNet data release 1.7, first remove those that cannot fit into any existing semantic frame in FrameNet; then, assign the remaining lexical units to their correct frames. We also present the Semi-supervised Deep Embedded Clustering with Anomaly Detection (SDEC-AD) model—an algorithm that maps high-dimensional contextualized vector representations of lexical units to a low-dimensional latent space for better frame prediction and uses reconstruction error to identify lexical units that cannot evoke frames in FrameNet. SDEC-AD outperforms the state-of-the-art methods in both steps of the frame induction process. Empirical results also show that definitions provide contextual information for representing and characterizing the frame membership of lexical units.

We propose a metric for machine translation evaluation based on frame semantics which does not require the use of reference translations or human corrections, but is aimed at comparing original and translated output directly. The metrics is described on the basis of an existing manual frame-semantic annotation of a parallel corpus with an English original and a Brazilian Portuguese and a German translation. We discuss implications of our metrics design, including the potential of scaling it for multiple languages.