BibTeX

@inProceedings{periti-etal-2024-automatically-343719,
	title        = {Automatically Generated Definitions and their utility for Modeling Word Meaning},
	abstract     = {Modeling lexical semantics is a challenging task, often suffering from interpretability pitfalls. In this paper, we delve into the generation of dictionary-like sense definitions and explore their utility for modeling word meaning. We fine-tuned two Llama models and include an existing T5-based model in our evaluation. Firstly, we evaluate the quality of the generated definitions on existing English benchmarks, setting new state-of-the-art results for the Definition Generation task. Next, we explore the use of definitions generated by our models as intermediate representations subsequently encoded as sentence embeddings. We evaluate this approach on lexical semantics tasks such as the Word-in-Context, Word Sense Induction, and Lexical Semantic Change, setting new state-of-the-art results in all three tasks when compared to unsupervised baselines.},
	booktitle    = {Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},
	author       = {Periti, Francesco and Alfter, David and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	pages        = {14008–14026},
}

@inProceedings{periti-etal-2024-trotr-343721,
	title        = {TRoTR: A Framework for Evaluating the Re-contextualization of Text Reuse
},
	abstract     = {Current approaches for detecting text reuse do not focus on recontextualization, i.e., how the new context(s) of a reused text differs from its original context(s). In this paper, we propose a novel framework called TRoTR that relies on the notion of topic relatedness for evaluating the diachronic change of context in which text is reused. TRoTR includes two NLP tasks: TRiC and TRaC. TRiC is designed to evaluate the topic relatedness between a pair of recontextualizations. TRaC is designed to evaluate the overall topic variation within a set of recontextualizations. We also provide a curated TRoTR benchmark of biblical text reuse, human-annotated with topic relatedness. The benchmark exhibits an inter-annotator agreement of .811. We evaluate multiple, established SBERT models on the TRoTR tasks and find that they exhibit greater sensitivity to textual similarity than topic relatedness. Our experiments show that fine-tuning these models can mitigate such a kind of sensitivity.},
	booktitle    = {Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},
	author       = {Periti, Francesco and Cassotti, Pierluigi and Montanelli, Stefano and Tahmasebi, Nina and Schlechtweg, Dominik},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	pages        = {13972–13990},
}

@inProceedings{schlechtweg-etal-2024-more-343718,
	title        = {More DWUGs: Extending and Evaluating Word Usage Graph Datasets in Multiple Languages},
	abstract     = {Word Usage Graphs (WUGs) represent human semantic proximity judgments for pairs of word uses in a weighted graph, which can be clustered to infer word sense clusters from simple pairwise word use judgments, avoiding the need for word sense definitions. SemEval-2020 Task 1 provided the first and to date largest manually annotated, diachronic WUG dataset. In this paper, we check the robustness and correctness of the annotations by continuing the SemEval annotation algorithm for two more rounds and comparing against an established annotation paradigm. Further, we test the reproducibility by resampling a new, smaller set of word uses from the SemEval source corpora and annotating them. Our work contributes to a better understanding of the problems and opportunities of the WUG annotation paradigm and points to future improvements.},
	booktitle    = {Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},
	author       = {Schlechtweg, Dominik and Cassotti, Pierluigi and Noble, Bill and Alfter, David and Schulte Im Walde, Sabine and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	pages        = {14379–14393},
}

@inProceedings{cassotti-etal-2024-using-338833,
	title        = {Using Synchronic Definitions and Semantic Relations to Classify Semantic Change Types},
	abstract     = {There is abundant evidence of the fact that the way words change their meaning can be classified in different types of change, highlighting the relationship between the old and new meanings (among which generalization, specialization and co-hyponymy transfer). In this paper, we present a way of detecting these types of change by constructing a model that leverages information both from synchronic lexical relations and definitions of word meanings. Specifically, we use synset definitions and hierarchy information from WordNet and test it on a digitized version of Blank's (1997) dataset of semantic change types. Finally, we show how the sense relationships can improve models for both approximation of human judgments of semantic relatedness as well as binary Lexical Semantic Change Detection. },
	booktitle    = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
	author       = {Cassotti, Pierluigi and De Pascale, Stefano and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
}

@inProceedings{noble-etal-2024-improving-339991,
	title        = {Improving Word Usage Graphs with Edge Induction},
	abstract     = {This paper investigates edge induction as a method for augmenting Word Usage Graphs, in which word usages (nodes) are connected through scores (edges) representing semantic relatedness. Clustering (densely) annotated WUGs can be used as a way to find senses of a word without relying on traditional word sense annotation. However, annotating all or a majority of pairs of usages is typically infeasible, resulting in sparse graphs and, likely, lower quality senses. In this paper, we ask if filling out WUGs with edges predicted from the human annotated edges improves the eventual clusters. We experiment with edge induction models that use structural features of the existing sparse graph, as well as those that exploit textual (distributional) features of the usages. We find that in both cases, inducing edges prior
to clustering improves correlation with human sense-usage annotation across three different clustering algorithms and languages.},
	booktitle    = {Proceedings of the 5th Workshop on Computational Approaches to Historical Language Change, August 15, 2024, Bangkok, Thailand},
	author       = {Noble, Bill and Periti, Francesco and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	address      = {Stroudsburg, PA},
	ISBN         = {979-8-89176-138-4},
}

@inProceedings{virk-etal-2024-enhancing-343103,
	title        = {Enhancing Swedish Parliamentary Data: Annotation, Accessibility, and Application in Digital Humanities},
	abstract     = {The Swedish bicameral parliament data presents a valuable textual resource that is of interest for many researches and scholars. The parliamentary texts offer many avenues for re- search including the study of how various af- fairs were run by governments over time. The Parliament proceedings are available in tex- tual format, but in their original form, they are noisy and unstructured and thus hard to explore and investigate. In this paper, we report the transformation of the raw bicameral parliament data (1867-1970) into a structured lexical re- source annotated with various word and doc- ument level attributes. The annotated data is then made searchable through two modern cor- pus infrastructure components which provide a wide array of corpus exploration, visualization, and comparison options. To demonstrate the practical utility of this resource, we present a case study examining the transformation of the concept of ’market’ over time from a tangible physical entity to an abstract idea.},
	booktitle    = {Association for Computational Linguistics (ACL)},
	author       = {Virk, Shafqat Mumtaz and Ohlsson, Claes and Björck, Henrik and Tahmasebi, Nina and Runefelt, Leif},
	year         = {2024},
}

@inProceedings{schlechtweg-etal-2024-more-343019,
	title        = {More DWUGs: Extending and Evaluating Word Usage Graph Datasets in Multiple Languages},
	abstract     = {Word Usage Graphs (WUGs) represent human semantic proximity judgments for pairs of word uses in a weighted graph, which can be clustered to infer word sense clusters from simple pairwise word use judgments, avoiding the need for word sense definitions. SemEval-2020 Task 1 provided the first and to date largest manually annotated, diachronic WUG dataset. In this paper, we check the robustness and correctness of the annotations by continuing the SemEval annotation algorithm for two more rounds and comparing against an established annotation paradigm. Further, we test the reproducibility by resampling a new, smaller set of word uses from the SemEval source corpora and annotating them. Our work contributes to a better understanding of the problems and opportunities of the WUG annotation paradigm and points to future improvements.},
	booktitle    = {    Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},
	author       = {Schlechtweg, Dominik and Cassotti, Pierluigi and Noble, Bill and Alfter, David and Schulte Im Walde, Sabine and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	address      = {Miami, Florida, USA},
	pages        = {14379–14393},
}

@inProceedings{periti-etal-2024-automatically-343018,
	title        = {Automatically Generated Definitions and their utility for Modeling Word Meaning},
	abstract     = {Modeling lexical semantics is a challenging task, often suffering from interpretability pitfalls. In this paper, we delve into the generation of dictionary-like sense definitions and explore their utility for modeling word meaning. We fine-tuned two Llama models and include an existing T5-based model in our evaluation. Firstly, we evaluate the quality of the generated definitions on existing English benchmarks, setting new state-of-the-art results for the Definition Generation task. Next, we explore the use of definitions generated by our models as intermediate representations subsequently encoded as sentence embeddings. We evaluate this approach on lexical semantics tasks such as the Word-in-Context, Word Sense Induction, and Lexical Semantic Change, setting new state-of-the-art results in all three tasks when compared to unsupervised baselines.},
	booktitle    = {Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},
	author       = {Periti, Francesco and Alfter, David and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	address      = {Miami, Florida, USA},
	pages        = {14008----14026},
}

@inProceedings{periti-etal-2024-trotr-343017,
	title        = {TRoTR: A Framework for Evaluating the Re-contextualization of Text Reuse},
	abstract     = {Current approaches for detecting text reuse do not focus on recontextualization, i.e., how the new context(s) of a reused text differs from its original context(s). In this paper, we propose a novel framework called TRoTR that relies on the notion of topic relatedness for evaluating the diachronic change of context in which text is reused. TRoTR includes two NLP tasks: TRiC and TRaC. TRiC is designed to evaluate the topic relatedness between a pair of recontextualizations. TRaC is designed to evaluate the overall topic variation within a set of recontextualizations. We also provide a curated TRoTR benchmark of biblical text reuse, human-annotated with topic relatedness. The benchmark exhibits an inter-annotator agreement of .811. We evaluate multiple, established SBERT models on the TRoTR tasks and find that they exhibit greater sensitivity to textual similarity than topic relatedness. Our experiments show that fine-tuning these models can mitigate such a kind of sensitivity.},
	booktitle    = {Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing},
	author       = {Periti, Francesco and Cassotti, Pierluigi and Montanelli, Stefano and Tahmasebi, Nina and Schlechtweg, Dominik},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	address      = {Miami, Florida, USA},
	pages        = {13972–13990},
}

@inProceedings{schlechtweg-etal-2024-durel-336715,
	title        = {The DURel Annotation Tool: Human and Computational Measurement of Semantic Proximity, Sense Clusters and Semantic Change},
	abstract     = {We present the DURel tool implementing the annotation of semantic proximity between word uses into an online, open source interface. The tool supports standardized human annotation as well as computational annotation, building on recent advances with Word-in-Context models. Annotator judgments are clustered with automatic graph clustering techniques and visualized for analysis. This allows to measure word senses with simple and intuitive micro-task judgments between use pairs, requiring minimal preparation efforts. The tool offers additional functionalities to compare the agreement between annotators to guarantee the inter-subjectivity of the obtained judgments and to calculate summary statistics over the annotated data giving insights into sense frequency distributions, semantic variation or changes of senses over time.},
	booktitle    = {Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, March 17-22, 2024, St. Julians, Malta. },
	author       = {Schlechtweg, Dominik and Virk, Shafqat and Sander, Pauline and Sköldberg, Emma and Theuer Linke, Lukas and Zhang, Tuo and Tahmasebi, Nina and Schulte im Walde, Sabine},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	ISBN         = {979-8-89176-091-2},
}

@inProceedings{periti-etal-2024-(chat)gpt-337358,
	title        = {(Chat)GPT v BERT Dawn of Justice for Semantic Change Detection},
	abstract     = {In the universe of Natural Language Processing, Transformer-based language models like BERT and (Chat)GPT have emerged as lexical superheroes with great power to solve open research problems. In this paper, we specifically focus on the temporal problem of semantic change, and evaluate their ability to solve two diachronic extensions of the Word-in-Context (WiC) task: TempoWiC and HistoWiC. In particular, we investigate the potential of a novel, off-the-shelf technology like ChatGPT (and GPT) 3.5 compared to BERT, which represents a family of models that currently stand as the state-of-the-art for modeling semantic change. Our experiments represent the first attempt to assess the use of (Chat)GPT for studying semantic change. Our results indicate that ChatGPT performs significantly worse than the foundational GPT version. Furthermore, our results demonstrate that (Chat)GPT achieves slightly lower performance than BERT in detecting long-term changes but performs significantly worse in detecting short-term changes.},
	booktitle    = {Findings of the Association for Computational Linguistics: EACL 2024, March 17-22, 2024, St. Julian’s, Malta},
	author       = {Periti, Francesco and Dubossarsky, Haim and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	ISBN         = {979-8-89176-093-6},
}

@article{periti-etal-2024-studying-340876,
	title        = {Studying word meaning evolution through incremental semantic shift detection},
	abstract     = {The study of semantic shift, that is, of how words change meaning as a consequence of social practices, events and political circumstances, is relevant in Natural Language Processing, Linguistics, and Social Sciences. The increasing availability of large diachronic corpora and advance in computational semantics have accelerated the development of computational approaches to detecting such shift. In this paper, we introduce a novel approach to tracing the evolution of word meaning over time. Our analysis focuses on gradual changes in word semantics and relies on an incremental approach to semantic shift detection (SSD) called What is Done is Done (WiDiD). WiDiD leverages scalable and evolutionary clustering of contextualised word embeddings to detect semantic shift and capture temporal transactions in word meanings. Existing approaches to SSD: (a) significantly simplify the semantic shift problem to cover change between two (or a few) time points, and (b) consider the existing corpora as static. We instead treat SSD as an organic process in which word meanings evolve across tens or even hundreds of time periods as the corpus is progressively made available. This results in an extremely demanding task that entails a multitude of intricate decisions. We demonstrate the applicability of this incremental approach on a diachronic corpus of Italian parliamentary speeches spanning eighteen distinct time periods. We also evaluate its performance on seven popular labelled benchmarks for SSD across multiple languages. Empirical results show that our results are comparable to state-of-the-art approaches, while outperforming the state-of-the-art for certain languages.},
	journal      = {Language Resources and Evaluation},
	author       = {Periti, Francesco and Picascia, Sergio and Montanelli, Stefano and Ferrara, Alfio and Tahmasebi, Nina},
	year         = {2024},
	pages        = {37},
}

@inProceedings{cassotti-etal-2024-computational-337360,
	title        = {Computational modeling of semantic change

},
	abstract     = {Languages change constantly over time, influenced by social, technological, cultural and political factors that affect how people express themselves. In particular, words can undergo the process of semantic change, which can be subtle and significantly impact the interpretation of texts. For example, the word terrific used to mean ‘causing terror’ and was as such synonymous to terrifying. Nowadays, speakers use the word in the sense of ‘excessive’ and even ‘amazing’. In Historical Linguistics, tools and methods have been developed to analyse this phenomenon, including systematic categorisations of the types of change, the causes and the mechanisms underlying the different types of change. However, traditional linguistic methods, while informative, are often based on small, carefully curated samples. Thanks to the availability of both large diachronic corpora, the computational means to model word meaning unsupervised, and evaluation benchmarks, we are seeing an increasing interest in the computational modelling of semantic change. This is evidenced by the increasing number of publications in this new domain as well as the organisation of initiatives and events related to this topic, such as four editions of the International Workshop on Computational Approaches to Historical Language Change LChange1, and several evaluation campaigns (Schlechtweg et al., 2020a; Basile et al., 2020b; Kutuzov et al.; Zamora-Reina et al., 2022).},
	booktitle    = {Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: Tutorial Abstracts},
	author       = {Cassotti, Pierluigi and Periti, Francesco and De Pascale, Stefano and Dubossarsky, Haim and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
}

@inProceedings{periti-tahmasebi-2024-towards-339990,
	title        = {Towards a Complete Solution to Lexical Semantic Change: an Extension to Multiple Time Periods and Diachronic Word Sense Induction},
	abstract     = {Thus far, the research community has focused on a simplified computational modeling of semantic change between two time periods. This simplified view has served as a foundational block but is not a complete solution to the complex modeling of semantic change. Acknowledging the power of recent language models, we believe that now is the right time to extend the current modeling to multiple time periods and diachronic word sense induction. In this position paper, we outline several extensions of
the current modeling and discuss issues related to the extensions. },
	booktitle    = {Proceedings of the 5th Workshop on Computational Approaches to Historical Language Change, Aug 15, 2024, Bangkok, Thailand},
	author       = {Periti, Francesco and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	address      = {Stroudsburg, PA},
	ISBN         = {979-8-89176-138-4},
}

@inProceedings{periti-etal-2024-analyzing-338831,
	title        = {Analyzing Semantic Change through Lexical Replacements},
	abstract     = {Modern language models are capable of contextualizing words based on their surrounding context. However, this capability is often compromised due to semantic change that leads to words being used in new, unexpected contexts not encountered during pre-training. In this paper, we model \textit{semantic change} by studying the effect of unexpected contexts introduced by lexical replacements. We propose a replacement schema where a target word is substituted with lexical replacements of varying relatedness, thus simulating different kinds of semantic change. Furthermore, we leverage the replacement schema as a basis for a novel interpretable model for semantic change. We are also the first to evaluate the use of LLaMa for semantic change detection. },
	booktitle    = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
	author       = {Periti, Francesco and Cassotti, Pierluigi and Dubossarsky, Haim and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
}

@inProceedings{periti-tahmasebi-2024-systematic-337365,
	title        = {A Systematic Comparison of Contextualized Word Embeddings for Lexical Semantic Change},
	abstract     = {Contextualized embeddings are the preferred tool for modeling Lexical Semantic Change (LSC). Current evaluations typically focus on a specific task known as Graded Change Detection (GCD). However, performance comparison across work are often misleading due to their reliance on diverse settings. In this paper, we evaluate state-of-the-art models and approaches for GCD under equal conditions. We further break the LSC problem into Word-in-Context (WiC) and Word Sense Induction (WSI) tasks, and compare models across these different levels. Our evaluation is performed across different languages on eight available benchmarks for LSC, and shows that (i) APD outperforms other approaches for GCD; (ii) XL-LEXEME outperforms other contextualized models for WiC, WSI, and GCD, while being comparable to GPT-4; (iii) there is a clear need for improving the modeling of word meanings, as well as focus on how, when, and why these meanings change, rather than solely focusing on the extent of semantic change.},
	booktitle    = {Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), June 16-21, 2024, Mexico City, Mexico},
	author       = {Periti, Francesco and Tahmasebi, Nina},
	year         = {2024},
	publisher    = {Association for Computational Linguistics},
	ISBN         = {979-8-89176-114-8},
}

@article{nielbo-etal-2024-quantitative-337356,
	title        = {Quantitative text analysis},
	abstract     = {Text analysis has undergone substantial evolution since its inception, moving from manual qualitative assessments to sophisticated quantitative and computational methods. Beginning in the late twentieth century, a surge in the utilization of computational techniques reshaped the landscape of text analysis, catalysed by advances in computational power and database technologies. Researchers in various fields, from history to medicine, are now using quantitative methodologies, particularly machine learning, to extract insights from massive textual data sets. This transformation can be described in three discernible methodological stages: feature-based models, representation learning models and generative models. Although sequential, these stages are complementary, each addressing analytical challenges in the text analysis. The progression from feature-based models that require manual feature engineering to contemporary generative models, such as GPT-4 and Llama2, signifies a change in the workflow, scale and computational infrastructure of the quantitative text analysis. This Primer presents a detailed introduction of some of these developments, offering insights into the methods, principles and applications pertinent to researchers embarking on the quantitative text analysis, especially within the field of machine learning.},
	journal      = {Nature Reviews Methods Primers},
	author       = {Nielbo, Kristoffer L. and Karsdorp, Folgert and Wevers, Melvin and Lassche, Alie and Baglini, Rebekah B. and Kestemont, Mike and Tahmasebi, Nina},
	year         = {2024},
	volume       = {4},
	number       = {1},
}