Eva Vanmassenhove

Javad Pourmostafa Roshan Sharami, Dimitar Shterionov, Frédéric Blain et al.

While quality estimation (QE) can play an important role in the translation process, its effectiveness relies on the availability and quality of training data. For QE in particular, high-quality labeled data is often lacking due to the high cost and effort associated with labeling such data. Aside from the data scarcity challenge, QE models should also be generalizable, i.e., they should be able to handle data from different domains, both generic and specific. To alleviate these two main issues -- data scarcity and domain mismatch -- this paper combines domain adaptation and data augmentation within a robust QE system. Our method first trains a generic QE model and then fine-tunes it on a specific domain while retaining generic knowledge. Our results show a significant improvement for all the language pairs investigated, better cross-lingual inference, and a superior performance in zero-shot learning scenarios as compared to state-of-the-art baselines.

Chiara Manna, Hosein Mohebbi, Afra Alishahi et al.

While Large Language Models achieve state-of-the-art results across a wide range of NLP tasks, they remain prone to systematic biases. Among these, gender bias is particularly salient in MT, due to systematic differences across languages in whether and how gender is marked. As a result, translation often requires disambiguating implicit source signals into explicit gender-marked forms. In this context, standard benchmarks may capture broad disparities but fail to reflect the full complexity of gender bias in modern MT. In this paper, we extend recent frameworks on bias evaluation by: (i) introducing a novel measure coined "Prior Bias", capturing a model's default gender assumptions, and (ii) applying the framework to decoder-only MT models. Our results show that, despite their scale and state-of-the-art status, decoder-only models do not generally outperform encoder-decoder architectures on gender-specific metrics; however, post-training (e.g., instruction tuning) not only improves contextual awareness but also reduces the masculine Prior Bias.

Argentina Anna Rescigno, Eva Vanmassenhove, Johanna Monti

Handling gender across languages remains a persistent challenge for Machine Translation (MT) and Large Language Models (LLMs), especially when translating from gender-neutral languages into morphologically gendered ones, such as English to Italian. English largely omits grammatical gender, while Italian requires explicit agreement across multiple grammatical categories. This asymmetry often leads MT systems to default to masculine forms, reinforcing bias and reducing translation accuracy. To address this issue, we present the Contextual Gender Annotation (ConGA) framework, a linguistically grounded set of guidelines for word-level gender annotation. The scheme distinguishes between semantic gender in English through three tags, Masculine (M), Feminine (F), and Ambiguous (A), and grammatical gender realisation in Italian (Masculine (M), Feminine (F)), combined with entity-level identifiers for cross-sentence tracking. We apply ConGA to the gENder-IT dataset, creating a gold-standard resource for evaluating gender bias in translation. Our results reveal systematic masculine overuse and inconsistent feminine realisation, highlighting persistent limitations of current MT systems. By combining fine-grained linguistic annotation with quantitative evaluation, this work offers both a methodology and a benchmark for building more gender-aware and multilingual NLP systems.

Jonas Klein, Chiara Manna, Eva Vanmassenhove

Gender bias in large language models has primarily been investigated for English, while languages with grammatical or morphological gender remain comparatively understudied. This paper investigates how and when gender information emerges in a Dutch BERT model trained from scratch, offering one of the first checkpoint-level analyses of bias formation in a Transformer architecture for a language combining overt morphological gender marking and generic forms. By extracting contextual embeddings throughout training, we construct dynamic gender subspaces using linear SVMs to trace when gender becomes linearly encoded and how this encoding evolves over time. Contextual embeddings are often assumed to integrate contextual cues robustly, allowing models to adjust the representation of a word depending on its more local usage. We therefore test whether explicit gender cues in controlled sentence templates (e.g., Zij is een loodgieter ('She is a plumber')) can override learned statistical associations (plumber -> male). Our findings challenge this assumption: although gender becomes clearly linearly separable around epoch 20 and is distributed across multiple embedding dimensions, the model struggles to update its internal gender representation in light of explicit contextual cues in short sentence templates. Stereotypical gender-profession pairings are predicted far more accurately than anti-stereotypical ones, and generic forms in Dutch systematically default to a male interpretation, even when the context explicitly denotes a female referent. Together, our results seem to indicate that contextualization in the representations learned by our Dutch BERT model is not sufficiently dynamic along the probed gender direction: explicit gender cues in anti-stereotypical contexts are not reliably reflected in the resulting representations, resulting in persistent male-default behaviour.

Çiçek Güven, Afra Alishahi, Henry Brighton et al.

In this paper, we elaborate on how AI can support diversity and inclusion and exemplify research projects conducted in that direction. We start by looking at the challenges and progress in making large language models (LLMs) more transparent, inclusive, and aware of social biases. Even though LLMs like ChatGPT have impressive abilities, they struggle to understand different cultural contexts and engage in meaningful, human like conversations. A key issue is that biases in language processing, especially in machine translation, can reinforce inequality. Tackling these biases requires a multidisciplinary approach to ensure AI promotes diversity, fairness, and inclusion. We also highlight AI's role in identifying biased content in media, which is important for improving representation. By detecting unequal portrayals of social groups, AI can help challenge stereotypes and create more inclusive technologies. Transparent AI algorithms, which clearly explain their decisions, are essential for building trust and reducing bias in AI systems. We also stress AI systems need diverse and inclusive training data. Projects like the Child Growth Monitor show how using a wide range of data can help address real world problems like malnutrition and poverty. We present a project that demonstrates how AI can be applied to monitor the role of search engines in spreading disinformation about the LGBTQ+ community. Moreover, we discuss the SignON project as an example of how technology can bridge communication gaps between hearing and deaf people, emphasizing the importance of collaboration and mutual trust in developing inclusive AI. Overall, with this paper, we advocate for AI systems that are not only effective but also socially responsible, promoting fair and inclusive interactions between humans and machines.

Chiara Manna, Afra Alishahi, Frédéric Blain et al.

While gender bias in modern Neural Machine Translation (NMT) systems has received much attention, traditional evaluation metrics do not to fully capture the extent to which these systems integrate contextual gender cues. We propose a novel evaluation metric called Minimal Pair Accuracy (MPA), which measures the reliance of models on gender cues for gender disambiguation. MPA is designed to go beyond surface-level gender accuracy metrics by focusing on whether models adapt to gender cues in minimal pairs -- sentence pairs that differ solely in the gendered pronoun, namely the explicit indicator of the target's entity gender in the source language (EN). We evaluate a number of NMT models on the English-Italian (EN--IT) language pair using this metric, we show that they ignore available gender cues in most cases in favor of (statistical) stereotypical gender interpretation. We further show that in anti-stereotypical cases, these models tend to more consistently take masculine gender cues into account while ignoring the feminine cues. Furthermore, we analyze the attention head weights in the encoder component and show that while all models encode gender information to some extent, masculine cues elicit a more diffused response compared to the more concentrated and specialized responses to feminine gender cues.

Eva Vanmassenhove

Multilingual Large Language Models (LLMs) considerably changed how technologies can influence language. While previous technologies could mediate or assist humans, there is now a tendency to offload the task of writing itself to these technologies, enabling them to change our linguistic ecosystem more directly. While they provide us quick access to information and impressively fluent output, beneath their apparent sophistication lies a subtle, more insidious threat: the gradual decline and loss of linguistic diversity. With this opinion piece, I explore how model collapse, with a particular focus on translation technology, can lead to the loss of linguistic forms, grammatical features, and cultural nuance. Model collapse refers to the eventual consequence of self-consuming training loops, where models reinforce their own biases and lose linguistic diversity. Drawing on recent work in Computer Vision, Natural Language Processing (NLP) and Machine Translation (MT), I argue that the tails of our linguistic distributions are vanishing, and with them, the narratives and identities they carry. This is a call to resist linguistic flattening and to reimagine NLP as a field that encourages, values and protects expressive multilingual lexical and linguistic diversity and creativity.

Eva Vanmassenhove

This chapter examines the role of Machine Translation in perpetuating gender bias, highlighting the challenges posed by cross-linguistic settings and statistical dependencies. A comprehensive overview of relevant existing work related to gender bias in both conventional Neural Machine Translation approaches and Generative Pretrained Transformer models employed as Machine Translation systems is provided. Through an experiment using ChatGPT (based on GPT-3.5) in an English-Italian translation context, we further assess ChatGPT's current capacity to address gender bias. The findings emphasize the ongoing need for advancements in mitigating bias in Machine Translation systems and underscore the importance of fostering fairness and inclusivity in language technologies.

Dimitar Shterionov, Eva Vanmassenhove

Over the past decade, deep learning (DL) has led to significant advancements in various fields of artificial intelligence, including machine translation (MT). These advancements would not be possible without the ever-growing volumes of data and the hardware that allows large DL models to be trained efficiently. Due to the large amount of computing cores as well as dedicated memory, graphics processing units (GPUs) are a more effective hardware solution for training and inference with DL models than central processing units (CPUs). However, the former is very power demanding. The electrical power consumption has economical as well as ecological implications. This chapter focuses on the ecological footprint of neural MT systems. It starts from the power drain during the training of and the inference with neural MT models and moves towards the environment impact, in terms of carbon dioxide emissions. Different architectures (RNN and Transformer) and different GPUs (consumer-grate NVidia 1080Ti and workstation-grade NVidia P100) are compared. Then, the overall CO2 offload is calculated for Ireland and the Netherlands. The NMT models and their ecological impact are compared to common household appliances to draw a more clear picture. The last part of this chapter analyses quantization, a technique for reducing the size and complexity of models, as a way to reduce power consumption. As quantized models can run on CPUs, they present a power-efficient inference solution without depending on a GPU.

Eva Vanmassenhove, Chris Emmery, Dimitar Shterionov

Recent years have seen an increasing need for gender-neutral and inclusive language. Within the field of NLP, there are various mono- and bilingual use cases where gender inclusive language is appropriate, if not preferred due to ambiguity or uncertainty in terms of the gender of referents. In this work, we present a rule-based and a neural approach to gender-neutral rewriting for English along with manually curated synthetic data (WinoBias+) and natural data (OpenSubtitles and Reddit) benchmarks. A detailed manual and automatic evaluation highlights how our NeuTral Rewriter, trained on data generated by the rule-based approach, obtains word error rates (WER) below 0.18% on synthetic, in-domain and out-domain test sets.

Eva Vanmassenhove, Johanna Monti

Languages differ in terms of the absence or presence of gender features, the number of gender classes and whether and where gender features are explicitly marked. These cross-linguistic differences can lead to ambiguities that are difficult to resolve, especially for sentence-level MT systems. The identification of ambiguity and its subsequent resolution is a challenging task for which currently there aren't any specific resources or challenge sets available. In this paper, we introduce gENder-IT, an English--Italian challenge set focusing on the resolution of natural gender phenomena by providing word-level gender tags on the English source side and multiple gender alternative translations, where needed, on the Italian target side.

Nishtha Jain, Maja Popovic, Declan Groves et al.

Gender bias is a frequent occurrence in NLP-based applications, especially pronounced in gender-inflected languages. Bias can appear through associations of certain adjectives and animate nouns with the natural gender of referents, but also due to unbalanced grammatical gender frequencies of inflected words. This type of bias becomes more evident in generating conversational utterances where gender is not specified within the sentence, because most current NLP applications still work on a sentence-level context. As a step towards more inclusive NLP, this paper proposes an automatic and generalisable rewriting approach for short conversational sentences. The rewriting method can be applied to sentences that, without extra-sentential context, have multiple equivalent alternatives in terms of gender. The method can be applied both for creating gender balanced outputs as well as for creating gender balanced training data. The proposed approach is based on a neural machine translation (NMT) system trained to 'translate' from one gender alternative to another. Both the automatic and manual analysis of the approach show promising results for automatic generation of gender alternatives for conversational sentences in Spanish.

Eva Vanmassenhove, Dimitar Shterionov, Matthew Gwilliam

Recent studies in the field of Machine Translation (MT) and Natural Language Processing (NLP) have shown that existing models amplify biases observed in the training data. The amplification of biases in language technology has mainly been examined with respect to specific phenomena, such as gender bias. In this work, we go beyond the study of gender in MT and investigate how bias amplification might affect language in a broader sense. We hypothesize that the 'algorithmic bias', i.e. an exacerbation of frequently observed patterns in combination with a loss of less frequent ones, not only exacerbates societal biases present in current datasets but could also lead to an artificially impoverished language: 'machine translationese'. We assess the linguistic richness (on a lexical and morphological level) of translations created by different data-driven MT paradigms - phrase-based statistical (PB-SMT) and neural MT (NMT). Our experiments show that there is a loss of lexical and morphological richness in the translations produced by all investigated MT paradigms for two language pairs (EN<=>FR and EN<=>ES).

Eva Vanmassenhove

New machine translations (MT) technologies are emerging rapidly and with them, bold claims of achieving human parity such as: (i) the results produced approach "accuracy achieved by average bilingual human translators" (Wu et al., 2017b) or (ii) the "translation quality is at human parity when compared to professional human translators" (Hassan et al., 2018) have seen the light of day (Laubli et al., 2018). Aside from the fact that many of these papers craft their own definition of human parity, these sensational claims are often not supported by a complete analysis of all aspects involved in translation. Establishing the discrepancies between the strengths of statistical approaches to MT and the way humans translate has been the starting point of our research. By looking at MT output and linguistic theory, we were able to identify some remaining issues. The problems range from simple number and gender agreement errors to more complex phenomena such as the correct translation of aspectual values and tenses. Our experiments confirm, along with other studies (Bentivogli et al., 2016), that neural MT has surpassed statistical MT in many aspects. However, some problems remain and others have emerged. We cover a series of problems related to the integration of specific linguistic features into statistical and neural MT, aiming to analyse and provide a solution to some of them. Our work focuses on addressing three main research questions that revolve around the complex relationship between linguistics and MT in general. We identify linguistic information that is lacking in order for automatic translation systems to produce more accurate translations and integrate additional features into the existing pipelines. We identify overgeneralization or 'algorithmic bias' as a potential drawback of neural MT and link it to many of the remaining linguistic issues.

Eva Vanmassenhove, Christian Hardmeier, Andy Way

Speakers of different languages must attend to and encode strikingly different aspects of the world in order to use their language correctly (Sapir, 1921; Slobin, 1996). One such difference is related to the way gender is expressed in a language. Saying "I am happy" in English, does not encode any additional knowledge of the speaker that uttered the sentence. However, many other languages do have grammatical gender systems and so such knowledge would be encoded. In order to correctly translate such a sentence into, say, French, the inherent gender information needs to be retained/recovered. The same sentence would become either "Je suis heureux", for a male speaker or "Je suis heureuse" for a female one. Apart from morphological agreement, demographic factors (gender, age, etc.) also influence our use of language in terms of word choices or even on the level of syntactic constructions (Tannen, 1991; Pennebaker et al., 2003). We integrate gender information into NMT systems. Our contribution is two-fold: (1) the compilation of large datasets with speaker information for 20 language pairs, and (2) a simple set of experiments that incorporate gender information into NMT for multiple language pairs. Our experiments show that adding a gender feature to an NMT system significantly improves the translation quality for some language pairs.

Eva Vanmassenhove, Dimitar Shterionov, Andy Way

This work presents an empirical approach to quantifying the loss of lexical richness in Machine Translation (MT) systems compared to Human Translation (HT). Our experiments show how current MT systems indeed fail to render the lexical diversity of human generated or translated text. The inability of MT systems to generate diverse outputs and its tendency to exacerbate already frequent patterns while ignoring less frequent ones, might be the underlying cause for, among others, the currently heavily debated issues related to gender biased output. Can we indeed, aside from biased data, talk about an algorithm that exacerbates seen biases?

Eva Vanmassenhove, Amit Moryossef, Alberto Poncelas et al.

We present our system for the CLIN29 shared task on cross-genre gender detection for Dutch. We experimented with a multitude of neural models (CNN, RNN, LSTM, etc.), more "traditional" models (SVM, RF, LogReg, etc.), different feature sets as well as data pre-processing. The final results suggested that using tokenized, non-lowercased data works best for most of the neural models, while a combination of word clusters, character trigrams and word lists showed to be most beneficial for the majority of the more "traditional" (that is, non-neural) models, beating features used in previous tasks such as n-grams, character n-grams, part-of-speech tags and combinations thereof. In contradiction with the results described in previous comparable shared tasks, our neural models performed better than our best traditional approaches with our best feature set-up. Our final model consisted of a weighted ensemble model combining the top 25 models. Our final model won both the in-domain gender prediction task and the cross-genre challenge, achieving an average accuracy of 64.93% on the in-domain gender prediction task, and 56.26% on cross-genre gender prediction.