Publications des membres de l’OLST

@inproceedings{zhao_worldvaluesbench_2024,

title = {WorldValuesBench: A Large-Scale Benchmark Dataset for Multi-Cultural Value Awareness of Language Models},

author = {Wenlong Zhao and Debanjan Mondal and Niket Tandon and Danica Dillion and Kurt Gray and Yuling Gu},

url = {https://aclanthology.org/2024.lrec-main.1539/},

year  = {2024},

date = {2024-01-01},

booktitle = {Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)},

pages = {17696–17706},

publisher = {ELRA and ICCL},

address = {Torino, Italy},

abstract = {The awareness of multi-cultural human values is critical to the ability of language models (LMs) to generate safe and personalized responses. However, this awareness of LMs has been insufficiently studied, since the computer science community lacks access to the large-scale real-world data about multi-cultural values. In this paper, we present WORLDVALUESBENCH, a globally diverse, large-scale benchmark dataset for the multi-cultural value prediction task, which requires a model to generate a rating response to a value question based on demographic contexts. Our dataset is derived from an influential social science project, World Values Survey (WVS), that has collected answers to hundreds of value questions (e.g., social, economic, ethical) from 94,728 participants worldwide. We have constructed more than 20 million examples of the type “(demographic attributes, value question) → answer” from the WVS responses. We perform a case study using our dataset and show that the task is challenging for strong open and closed-source models. On merely 11.1%, 25.0%, 72.2%, and 75.0% of the questions, Alpaca-7B, Vicuna-7B-v1.5, Mixtral-8x7B-Instruct-v0.1, and GPT-3.5 Turbo can respectively achieve < 0.2 Wasserstein 1-distance from the human normalized answer distributions. WORLDVALUESBENCH opens up new research avenues in studying limitations and opportunities in multi-cultural value awareness of LMs.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{zheng_large_2024,

title = {Large Language Models Are Not Robust Multiple Choice Selectors},

author = {Chujie Zheng and Hao Zhou and Fandong Meng and Jie Zhou and Minlie Huang},

year  = {2024},

date = {2024-01-01},

booktitle = {Proceedings of the Twelfth International Conference on Learning Representations},

address = {Vienna, Austria},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{yao_benchmarking_2024,

title = {Benchmarking Machine Translation with Cultural Awareness},

author = {Binwei Yao and Ming Jiang and Tara Bobinac and Diyi Yang and Junjie Hu},

url = {https://aclanthology.org/2024.findings-emnlp.765},

doi = {10.18653/v1/2024.findings-emnlp.765},

year  = {2024},

date = {2024-01-01},

urldate = {2025-03-06},

booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2024},

pages = {13078–13096},

publisher = {Association for Computational Linguistics},

address = {Miami, Florida, USA},

abstract = {Translating culture-related content is vital for effective cross-cultural communication. However, many culture-specific items (CSIs) often lack viable translations across languages, making it challenging to collect high-quality, diverse parallel corpora with CSI annotations. This difficulty hinders the analysis of cultural awareness of machine translation (MT) systems, including traditional neural MT and the emerging MT paradigm using large language models (LLM). To address this gap, we introduce a novel parallel corpus, enriched with CSI annotations in 6 language pairs for investigating Culturally-Aware Machine Translation—CAMT.1 Furthermore, we design two evaluation metrics to assess CSI translations, focusing on their pragmatic translation quality. Our findings show the superior ability of LLMs over neural MTs in leveraging external cultural knowledge for translating CSIs, especially those lacking translations in the target culture.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{rottger_political_2024,

title = {Political Compass or Spinning Arrow? Towards More Meaningful Evaluations for Values and Opinions in Large Language Models},

author = {Paul Röttger and Valentin Hofmann and Valentina Pyatkin and Musashi Hinck and Hannah Kirk and Hinrich Schuetze and Dirk Hovy},

url = {https://aclanthology.org/2024.acl-long.816},

doi = {10.18653/v1/2024.acl-long.816},

year  = {2024},

date = {2024-01-01},

urldate = {2024-10-28},

booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},

pages = {15295–15311},

publisher = {Association for Computational Linguistics},

address = {Bangkok, Thailand},

abstract = {Much recent work seeks to evaluate values and opinions in large language models (LLMs) using multiple-choice surveys and questionnaires. Most of this work is motivated by concerns around real-world LLM applications. For example, politically-biased LLMs may subtly influence society when they are used by millions of people. Such real-world concerns, however, stand in stark contrast to the artificiality of current evaluations: real users do not typically ask LLMs survey questions. Motivated by this discrepancy, we challenge the prevailing constrained evaluation paradigm for values and opinions in LLMs and explore more realistic unconstrained evaluations. As a case study, we focus on the popular Political Compass Test (PCT). In a systematic review, we find that most prior work using the PCT forces models to comply with the PCT’s multiple-choice format. We show that models give substantively different answers when not forced; that answers change depending on how models are forced; and that answers lack paraphrase robustness. Then, we demonstrate that models give different answers yet again in a more realistic open-ended answer setting. We distill these findings into recommendations and open challenges in evaluating values and opinions in LLMs.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{ren_valuebench_2024,

title = {ValueBench: Towards Comprehensively Evaluating Value Orientations and Understanding of Large Language Models},

author = {Yuanyi Ren and Haoran Ye and Hanjun Fang and Xin Zhang and Guojie Song},

url = {https://aclanthology.org/2024.acl-long.111},

doi = {10.18653/v1/2024.acl-long.111},

year  = {2024},

date = {2024-01-01},

urldate = {2024-10-28},

booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},

pages = {2015–2040},

publisher = {Association for Computational Linguistics},

address = {Bangkok, Thailand},

abstract = {Large Language Models (LLMs) are transforming diverse fields and gaining increasing influence as human proxies. This development underscores the urgent need for evaluating value orientations and understanding of LLMs to ensure their responsible integration into public-facing applications. This work introduces ValueBench, the first comprehensive psychometric benchmark for evaluating value orientations and value understanding in LLMs. ValueBench collects data from 44 established psychometric inventories, encompassing 453 multifaceted value dimensions. We propose an evaluation pipeline grounded in realistic human-AI interactions to probe value orientations, along with novel tasks for evaluating value understanding in an open-ended value space. With extensive experiments conducted on six representative LLMs, we unveil their shared and distinctive value orientations and exhibit their ability to approximate expert conclusions in value-related extraction and generation tasks. ValueBench is openly accessible at https://github.com/Value4AI/ValueBench.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{ryan_unintended_2024,

title = {Unintended Impacts of LLM Alignment on Global Representation},

author = {Michael J Ryan and William Held and Diyi Yang},

url = {https://aclanthology.org/2024.acl-long.853},

doi = {10.18653/v1/2024.acl-long.853},

year  = {2024},

date = {2024-01-01},

urldate = {2025-03-05},

booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},

pages = {16121–16140},

publisher = {Association for Computational Linguistics},

address = {Bangkok, Thailand},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{richburg_how_2024,

title = {How Multilingual Are Large Language Models Fine-Tuned for Translation?},

author = {Aquia Richburg and Marine Carpuat},

url = {https://openreview.net/forum?id=bnscREWUuc},

year  = {2024},

date = {2024-01-01},

booktitle = {Proceedings of the First Conference on Language Modelling},

address = {Philadelphia, Pennsylvania, United States},

abstract = {A new paradigm for machine translation has recently emerged: fine-tuning large language models (LLMs) on parallel text has been shown to outperform dedicated translation systems trained in a supervised fashion on much larger amounts of parallel data (Xu et al., 2024a; Alves et al., 2024). However, it remains unclear whether this paradigm can enable massively multilingual machine translation or whether it requires fine-tuning dedicated models for a small number of language pairs. How does translation fine-tuning impact the MT capabilities of LLMs for zero-shot languages, zero-shot language pairs, and translation tasks that do not involve English? To address these questions, we conduct an extensive empirical evaluation of the translation quality of the TOWER family of language models (Alves et al., 2024) on 132 translation tasks from the multi-parallel FLORES-200 dataset. We find that translation fine-tuning improves translation quality even for zero-shot languages on average, but that the impact is uneven depending on the language pairs involved. These results call for further research to effectively enable massively multilingual translation with LLMs.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{moore_are_2024,

title = {Are Large Language Models Consistent over Value-laden Questions?},

author = {Jared Moore and Tanvi Deshpande and Diyi Yang},

url = {https://aclanthology.org/2024.findings-emnlp.891},

doi = {10.18653/v1/2024.findings-emnlp.891},

year  = {2024},

date = {2024-01-01},

urldate = {2025-02-27},

booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2024},

pages = {15185–15221},

publisher = {Association for Computational Linguistics},

address = {Miami, Florida, USA},

abstract = {Large language models (LLMs) appear to bias their survey answers toward certain values. Nonetheless, some argue that LLMs are too inconsistent to simulate particular values. Are they? To answer, we first define value consistency as the similarity of answers across (1) paraphrases of one question, (2) related questions under one topic, (3) multiple-choice and open-ended use-cases of one question, and (4) multilingual translations of a question to English, Chinese, German, and Japanese. We apply these measures to small and large, open LLMs including llama-3, as well as gpt-4o, using 8,000 questions spanning more than 300 topics. Unlike prior work, we find that models are relatively consistent across paraphrases, use-cases, translations, and within a topic. Still, some inconsistencies remain. Models are more consistent on uncontroversial topics (e.g., in the U.S., "Thanksgiving") than on controversial ones ("euthanasia"). Base models are both more consistent compared to fine-tuned models and are uniform in their consistency across topics, while fine-tuned models are more inconsistent about some topics ("euthanasia") than others ("women’s rights") like our human subjects (n=165).},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{mihaylov_what_2024,

title = {What an Elegant Bridge: Multilingual LLMs are Biased Similarly in Different Languages},

author = {Viktor Mihaylov and Aleksandar Shtedritski},

url = {https://aclanthology.org/2024.nlp4science-1.3/},

doi = {10.18653/v1/2024.nlp4science-1.3},

year  = {2024},

date = {2024-01-01},

booktitle = {Proceedings of the 1st Workshop on NLP for Science (NLP4Science)},

pages = {16–23},

publisher = {Association for Computational Linguistics},

address = {Miami, FL, USA},

abstract = {This paper investigates biases of Large Language Models (LLMs) through the lens of grammatical gender. Drawing inspiration from seminal works in psycholinguistics, particularly the study of gender’s influence on language perception, we leverage multilingual LLMs to revisit and expand upon the foundational experiments of Boroditsky (2003). Employing LLMs as a novel method for examining psycholinguistic biases related to grammatical gender, we prompt a model to describe nouns with adjectives in various languages, focusing specifically on languages with grammatical gender. In particular, we look at adjective co-occurrences across gender and languages, and train a binary classifier to predict grammatical gender given adjectives an LLM uses to describe a noun. Surprisingly, we find that a simple classifier can not only predict noun gender above chance but also exhibit crosslanguage transferability. We find a strong social influence of language on the way multilingual LLMs reason.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{masoud_cultural_2024,

title = {Cultural Alignment in Large Language Models: An Explanatory Analysis Based on Hofstede's Cultural Dimensions},

author = {Reem Masoud and Ziquan Liu and Martin Ferianc and Philip C Treleaven and Miguel Rodrigues Rodrigues},

year  = {2024},

date = {2024-01-01},

booktitle = {Proceedings of the 31st International Conference on Computational Linguistics},

pages = {8474–8503},

publisher = {Association for Computational Linguistics},

address = {Abu Dhabi, UAE},

abstract = {The deployment of large language models 

(LLMs) raises concerns regarding their cultural misalignment and potential ramifications 

on individuals and societies with diverse cultural backgrounds. While the discourse has 

focused mainly on political and social biases, our research proposes a Cultural Alignment Test (Hoftede’s CAT) to quantify cultural 

alignment using Hofstede’s cultural dimension 

framework, which offers an explanatory crosscultural comparison through the latent variable 

analysis. We apply our approach to quantitatively evaluate LLMs—namely Llama 2, GPT3.5, and GPT-4—against the cultural dimensions of regions like the United States, China, 

and Arab countries, using different prompting 

styles and exploring the effects of languagespecific fine-tuning on the models’ behavioural 

tendencies and cultural values. Our results 

quantify the cultural alignment of LLMs and 

reveal the difference between LLMs in explanatory cultural dimensions. Our study demonstrates that while all LLMs struggle to grasp 

cultural values, GPT-4 shows a unique capability to adapt to cultural nuances, particularly in Chinese settings. However, it faces 

challenges with American and Arab cultures. 

The research also highlights that fine-tuning 

LLama 2 models with different languages 

changes their responses to cultural questions, 

emphasizing the need for culturally diverse 

development in AI for worldwide acceptance 

and ethical use. For more details or to contribute to this research, visit our GitHub page 

https://github.com/reemim/Hofstedes_CAT.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{mukherjee_cultural_2024,

title = {Cultural Conditioning or Placebo? On the Effectiveness of Socio-Demographic Prompting},

author = {Sagnik Mukherjee and Muhammad Farid Adilazuarda and Sunayana Sitaram and Kalika Bali and Alham Fikri Aji and Monojit Choudhury},

url = {https://aclanthology.org/2024.emnlp-main.884},

doi = {10.18653/v1/2024.emnlp-main.884},

year  = {2024},

date = {2024-01-01},

urldate = {2025-03-01},

booktitle = {Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},

pages = {15811–15837},

publisher = {Association for Computational Linguistics},

address = {Miami, Florida, USA},

abstract = {Socio-demographic prompting is a commonly employed approach to study cultural biases in LLMs as well as for aligning models to certain cultures. In this paper, we systematically probe four LLMs (Llama 3, Mistral v0.2, GPT3.5 Turbo and GPT-4) with prompts that are conditioned on culturally sensitive and nonsensitive cues, on datasets that are supposed to be culturally sensitive (EtiCor and CALI) or neutral (MMLU and ETHICS). We observe that all models except GPT-4 show significant variations in their responses on both kinds of datasets for both kinds of prompts, casting doubt on the robustness of the culturallyconditioned prompting as a method for eliciting cultural bias in models or as an alignment strategy. The work also calls rethinking the control experiment design to tease apart the cultural conditioning of responses from “placebo effect”, i.e., random perturbations of model responses due to arbitrary tokens in the prompt.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{benkler_assessing_2023,

title = {Assessing LLMs for Moral Value Pluralism},

author = {Noam Benkler and Drisana Mosaphir and Scott Friedman and Andrew Smart and Sonja Schmer-Galunder},

url = {http://arxiv.org/abs/2312.10075},

doi = {10.48550/arXiv.2312.10075},

year  = {2023},

date = {2023-12-01},

urldate = {2025-02-28},

booktitle = {Proceedings of the 37th Conference on Neural Information Processing Systems (NeurIPS 2023)},

publisher = {arXiv},

address = {New Orleans, Louisiana, USA},

abstract = {Moral values are important indicators of socio-cultural norms and behavior and guide our moral judgment and identity. Decades of social science research have developed and refined some widely-accepted surveys, such as the World Values Survey (WVS), that elicit value judgments from direct questions, enabling social scientists to measure higher-level moral values and even cultural value distance. While WVS is accepted as an explicit assessment of values, we lack methods for assessing implicit moral and cultural values in media, e.g., encountered in social media, political rhetoric, narratives, and generated by AI systems such as the large language models (LLMs) that are taking foothold in our daily lives. As we consume online content and utilize LLM outputs, we might ask, practically or academically, which moral values are being implicitly promoted or undercut, or—in the case of LLMs—if they are intending to represent a cultural identity, are they doing so consistently? In this paper we utilize a Recognizing Value Resonance (RVR) NLP model to identify WVS values that resonate and conflict with a passage of text. We apply RVR to the text generated by LLMs to characterize implicit moral values, allowing us to quantify the moral/cultural distance between LLMs and various demographics that have been surveyed using the WVS. Our results highlight value misalignment for non-WEIRD nations from various clusters of the WVS cultural map, as well as age misalignment across nations.},

note = {arXiv:2312.10075 [cs]},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{marshman_updating_2023,

title = {Updating translator education programs: Adapting to technologies and their impacts in the Canadian language industry},

author = {Elizabeth Marshman and Anwar Ghanim Alfetlawi and Haifa Ben Naji and Dipen Dave and Ahmed Elhuseiny Bedeir and Ting Liu},

year  = {2023},

date = {2023-11-01},

publisher = {ASLING},

address = {Luxembourg},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{marshman_updating_2023-1,

title = {Updating TTI programs for today’s Canadian market: A case study at uOttawa},

author = {Elizabeth Marshman and Anwar Ghanim Alfetlawi and Haifa Ben Naji and Dipen Dave and Ahmed Elhuseiny Bedeir and Ting Liu},

year  = {2023},

date = {2023-10-01},

address = {Monterey, California/Virtual},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{mille_mod-d2t_2023,

title = {Mod-D2T: A Multi-layer Dataset for Modular Data-to-Text Generation},

author = {Simon Mille and François Lareau and Stamatia Dasiopoulou and Anya Belz},

editor = {C. Maria Keet and Hung-Yi Lee and Sina Zarrieß},

url = {https://aclanthology.org/2023.inlg-main.36},

doi = {10.18653/v1/2023.inlg-main.36},

year  = {2023},

date = {2023-09-01},

urldate = {2024-05-31},

booktitle = {Proceedings of the 16th International Natural Language Generation Conference},

pages = {455–466},

publisher = {Association for Computational Linguistics},

address = {Prague, Czechia},

abstract = {Rule-based text generators lack the coverage and fluency of their neural counterparts, but have two big advantages over them: (i) they are entirely controllable and do not hallucinate; and (ii) they can fully explain how an output was generated from an input. In this paper we leverage these two advantages to create large and reliable synthetic datasets with multiple human-intelligible intermediate representations. We present the Modular Data-to-Text (Mod-D2T) Dataset which incorporates ten intermediate-level representations between input triple sets and output text; the mappings from one level to the next can broadly be interpreted as the traditional modular tasks of an NLG pipeline. We describe the Mod-D2T dataset, evaluate its quality via manual validation and discuss its applications and limitations. Data, code and documentation are available at https://github.com/mille-s/Mod-D2T.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{venant_predicates_2023,

title = {Predicates and entities in Abstract Meaning Representation},

author = {Antoine Venant and François Lareau},

editor = {Owen Rambow and François Lareau},

url = {https://aclanthology.org/2023.depling-1.4},

year  = {2023},

date = {2023-03-01},

urldate = {2024-05-31},

booktitle = {Proceedings of the Seventh International Conference on Dependency Linguistics (Depling, GURT/SyntaxFest 2023)},

pages = {32–41},

publisher = {Association for Computational Linguistics},

address = {Washington, D.C.},

abstract = {Nodes in Abstract Meaning Representation (AMR) are generally thought of as neo-Davidsonian entities. We review existing translation into neo-Davidsonian representations and show that these translations inconsistently handle copula sentences. We link the problem to an asymmetry arising from a problematic handling of words with no associated PropBank frames for the underlying predicate. We introduce a method to automatically and uniformly decompose AMR nodes into an entity-part and a predicative part, which offers a consistent treatment of copula sentences and quasi- predicates such as brother or client.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{keleg_dlama_2023,

title = {DLAMA: A Framework for Curating Culturally Diverse Facts for Probing the Knowledge of Pretrained Language Models},

author = {Amr Keleg and Walid Magdy},

url = {https://aclanthology.org/2023.findings-acl.389},

doi = {10.18653/v1/2023.findings-acl.389},

year  = {2023},

date = {2023-01-01},

urldate = {2025-02-22},

booktitle = {Findings of the Association for Computational Linguistics: ACL 2023},

pages = {6245–6266},

publisher = {Association for Computational Linguistics},

address = {Toronto, Canada},

abstract = {A few benchmarking datasets have been released to evaluate the factual knowledge of pretrained language models. These benchmarks (e.g., LAMA, and ParaRel) are mainly developed in English and later are translated to form new multilingual versions (e.g., mLAMA, and mParaRel). Results on these multilingual benchmarks suggest that using English prompts to recall the facts from multilingual models usually yields significantly better and more consistent performance than using non-English prompts. Our analysis shows that mLAMA is biased toward facts from Western countries, which might affect the fairness of probing models. We propose a new framework for curating factual triples from Wikidata that are culturally diverse. A new benchmark DLAMA-v1 is built of factual triples from three pairs of contrasting cultures having a total of 78,259 triples from 20 relation predicates. The three pairs comprise facts representing the (Arab and Western), (Asian and Western), and (South American and Western) countries respectively. Having a more balanced benchmark (DLAMA-v1) supports that mBERT performs better on Western facts than non-Western ones, while monolingual Arabic, English, and Korean models tend to perform better on their culturally proximate facts. Moreover, both monolingual and multilingual models tend to make a prediction that is culturally or geographically relevant to the correct label, even if the prediction is wrong.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{cao_assessing_2023,

title = {Assessing Cross-Cultural Alignment between ChatGPT and Human Societies: An Empirical Study},

author = {Yong Cao and Li Zhou and Seolhwa Lee and Laura Cabello and Min Chen and Daniel Hershcovich},

doi = {10.18653/v1/2023.c3nlp-1.7},

year  = {2023},

date = {2023-01-01},

booktitle = {Proceedings of the First Workshop on Cross-Cultural Considerations in NLP (C3NLP)},

pages = {53–67},

publisher = {Association for Computational Linguistics},

address = {Dubrovnik, Coratia},

abstract = {The recent release of ChatGPT has garnered widespread recognition for its exceptional ability to generate human-like responses in dialogue. Given its usage by users from various nations and its training on a vast multilingual corpus that incorporates diverse cultural and societal norms, it is crucial to evaluate its effectiveness in cultural adaptation. In this paper, we investigate the underlying cultural background of ChatGPT by analyzing its responses to questions designed to quantify human cultural differences. Our findings suggest that, when prompted with American context, ChatGPT exhibits a strong alignment with American culture, but it adapts less effectively to other cultural contexts. Furthermore, by using different prompts to probe the model, we show that English prompts reduce the variance in model responses, flattening out cultural differences and biasing them towards American culture. This study provides valuable insights into the cultural implications of ChatGPT and highlights the necessity of greater diversity and cultural awareness in language technologies.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{san_martin_extracting_2023,

title = {Extracting the Agent-Patient Relation from Corpus With Word Sketches},

author = {San Martín, Antonio and Trekker, Catherine and Díaz-Bautista, Juan Carlos},

url = {https://aclanthology.org/2023.ldk-1.73.pdf},

year  = {2023},

date = {2023-01-01},

booktitle = {Proceedings of the 4th Conference on Language, Data and Knowledge},

pages = {666--675},

publisher = {NOVA CLUNL},

address = {Vienna, Austria},

abstract = {Word sketches are a powerful function of Sketch Engine that automatically summarizes the most common usage patterns of a search word in a corpus. While they have proven to be a valuable tool for collocational analysis in both general and specialized language, their potential for the extraction of terminological knowledge is yet to be fully realized. To address this, we introduce a novel semantic sketch grammar designed to extract the agent-patient relation, an important yet understudied relation. This paper presents the various stages of developing the rules that compose this sketch grammar as well as the evaluation of their precision. The errors identified during the evaluation process are also analyzed to guide future improvements. The sketch grammar is available online so that any user can apply it to their own corpora in Sketch Engine.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{arora_probing_2023,

title = {Probing Pre-Trained Language Models for Cross-Cultural Differences in Values},

author = {Arnav Arora and Lucie-Aimée Kaffee and Isabelle Augenstein},

doi = {10.18653/v1/2023.c3nlp-1.12},

year  = {2023},

date = {2023-01-01},

booktitle = {Proceedings of the First Workshop on Cross-Cultural Considerations in NLP (C3NLP)},

pages = {114–130},

publisher = {Association for Computational Linguistic},

address = {Dubrovnik, Coratia},

abstract = {Language embeds information about social, cultural, and political values people hold. Prior work has explored potentially harmful social biases encoded in Pre-trained Language Models (PLMs). However, there has been no systematic study investigating how values embedded in these models vary across cultures. In this paper, we introduce probes to study which crosscultural values are embedded in these models, and whether they align with existing theories and cross-cultural values surveys. We find that PLMs capture differences in values across cultures, but those only weakly align with established values surveys. We discuss implications of using mis-aligned models in cross-cultural settings, as well as ways of aligning PLMs with values surveys.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

@inproceedings{rigouts_terryn_supervised_2023,

title = {Supervised Feature-based Classification Approach to Bilingual Lexicon Induction from Specialised Comparable Corpora},

author = {Ayla Rigouts Terryn},

doi = {https://doi.org/10.26615/978-954-452-090-8_008},

year  = {2023},

date = {2023-01-01},

booktitle = {Proceedings of the 2023 Workshop on Computational Terminology in NLP and Translation Studies (ConTeNTS), incorporating the BUCC workshop},

pages = {59–68},

publisher = {INCOMA Ltd. Shoumen},

address = {Varna, Bulgaria},

abstract = {This study, submitted to the BUCC2023 shared task on bilingual term alignment in comparable specialised corpora, introduces a supervised, feature-based classification approach. The approach employs both static cross-lingual embeddings and contextual multilingual embeddings, combined with surface-level indicators such as Levenshtein distance and term length, as well as linguistic information. Results exhibit improved performance over previous methodologies, illustrating the merit of integrating diverse features. However, the error analysis also reveals remaining challenges.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}