In this work, we explore whether the recently demonstrated zero-shot abilities of the T0 model extend to Named Entity Recognition for out-of-distribution languages and time periods. Using a historical newspaper corpus in 3 languages as test-bed, we use prompts to extract possible named entities. Our results show that a naive approach for prompt-based zero-shot multilingual Named Entity Recognition is error-prone, but highlights the potential of such an approach for historical languages lacking labeled datasets. Moreover, we also find that T0-like models can be probed to predict the publication date and language of a document, which could be very relevant for the study of historical texts.

Recent approaches to Word Sense Disambiguation (WSD) have profited from the enhanced contextualized word representations coming from contemporary Large Language Models (LLMs). This advancement is accompanied by a renewed interest in WSD applications in Humanities research, where the lack of suitable, specific WSD-annotated resources is a hurdle in developing ad-hoc WSD systems. Because they can exploit sentential context, LLMs are particularly suited for disambiguation tasks. Still, the application of LLMs is often limited to linear classifiers trained on top of the LLM architecture. In this paper, we follow recent developments in non-parametric learning and show how LLMs can be efficiently fine-tuned to achieve strong few-shot performance on WSD for historical languages (English and Dutch, date range: 1450-1950). We test our hypothesis using (i) a large, general evaluation set taken from large lexical databases, and (ii) a small real-world scenario involving an ad-hoc WSD task. Moreover, this paper marks the release of GysBERT, a LLM for historical Dutch.

We report on an inter-annotator agreement experiment involving instances of text reuse focusing on the well-known case of biblical intertextuality in medieval literature. We target the application use case of literary scholars whose aim is to document instances of biblical references in the ‘apparatus fontium’ of a prospective digital edition. We develop a Bayesian implementation of Cohen’s kappa for multiple annotators that allows us to assess the influence of various contextual effects on the inter-annotator agreement, producing both more robust estimates of the agreement indices as well as insights into the annotation process that leads to the estimated indices. As a result, we are able to produce a novel and nuanced estimation of inter-annotator agreement in the context of intertextuality, exploring the challenges that arise from manually annotating a dataset of biblical references in the writings of Bernard of Clairvaux. Among others, our method was able to unveil the fact that the obtained agreement depends heavily on the biblical source book of the proposed reference, as well as the underlying algorithm used to retrieve the candidate match.

The new pre-train-then-fine-tune paradigm in Natural made important performance gains accessible to a wider audience. Once pre-trained, deploying a large language model presents comparatively small infrastructure requirements, and offers robust performance in many NLP tasks. The Digital Humanities community has been an early adapter of this paradigm. Yet, a large part of this community is concerned with the application of NLP algorithms to historical texts, for which large models pre-trained on contemporary text may not provide optimal results. In the present paper, we present “MacBERTh”—a transformer-based language model pre-trained on historical English—and exhaustively assess its benefits on a large set of relevant downstream tasks. Our experiments highlight that, despite some differences across target time periods, pre-training on historical language from scratch outperforms models pre-trained on present-day language and later adapted to historical language.

The detection of allusive text reuse is particularly challenging due to the sparse evidence on which allusive references rely — commonly based on none or very few shared words. Arguably, lexical semantics can be resorted to since uncovering semantic relations between words has the potential to increase the support underlying the allusion and alleviate the lexical sparsity. A further obstacle is the lack of evaluation benchmark corpora, largely due to the highly interpretative character of the annotation process. In the present paper, we aim to elucidate the feasibility of automated allusion detection. We approach the matter from an Information Retrieval perspective in which referencing texts act as queries and referenced texts as relevant documents to be retrieved, and estimate the difficulty of benchmark corpus compilation by a novel inter-annotator agreement study on query segmentation. Furthermore, we investigate to what extent the integration of lexical semantic information derived from distributional models and ontologies can aid retrieving cases of allusive reuse. The results show that (i) despite low agreement scores, using manual queries considerably improves retrieval performance with respect to a windowing approach, and that (ii) retrieval performance can be moderately boosted with distributional semantics.

This paper addresses Hip-Hop lyric generation with conditional Neural Language Models. We develop a simple yet effective mechanism to extract and apply conditional templates from text snippets, and show—on the basis of a large-scale crowd-sourced manual evaluation—that these templates significantly improve the quality and realism of the generated snippets. Importantly, the proposed approach enables end-to-end training, targeting formal properties of text such as rhythm and rhyme, which are central characteristics of rap texts. Additionally, we explore how generating text at different scales (e.g. character-level or word-level) affects the quality of the output. We find that a hybrid form—a hierarchical model that aims to integrate Language Modeling at both word and character-level scales—yields significant improvements in text quality, yet surprisingly, cannot exploit conditional templates to their fullest extent. Our findings highlight that text generation models based on Recurrent Neural Networks (RNN) are sensitive to the modeling scale and call for further research on the observed differences in effectiveness of the conditioning mechanism at different scales.

Lemmatization of standard languages is concerned with (i) abstracting over morphological differences and (ii) resolving token-lemma ambiguities of inflected words in order to map them to a dictionary headword. In the present paper we aim to improve lemmatization performance on a set of non-standard historical languages in which the difficulty is increased by an additional aspect (iii): spelling variation due to lacking orthographic standards. We approach lemmatization as a string-transduction task with an Encoder-Decoder architecture which we enrich with sentence information using a hierarchical sentence encoder. We show significant improvements over the state-of-the-art by fine-tuning the sentence encodings to jointly optimize a bidirectional language model loss. Crucially, our architecture does not require POS or morphological annotations, which are not always available for historical corpora. Additionally, we also test the proposed model on a set of typologically diverse standard languages showing results on par or better than a model without fine-tuned sentence representations and previous state-of-the-art systems. Finally, to encourage future work on processing of non-standard varieties, we release the dataset of non-standard languages underlying the present study, which is based on openly accessible sources.

The task of obfuscating writing style using sequence models has previously been investigated under the framework of obfuscation-by-transfer, where the input text is explicitly rewritten in another style. A side effect of this framework are the frequent major alterations to the semantic content of the input. In this work, we propose obfuscation-by-invariance, and investigate to what extent models trained to be explicitly style-invariant preserve semantics. We evaluate our architectures in parallel and non-parallel settings, and compare automatic and human evaluations on the obfuscated sentences. Our experiments show that the performance of a style classifier can be reduced to chance level, while the output is evaluated to be of equal quality to models applying style-transfer. Additionally, human evaluation indicates a trade-off between the level of obfuscation and the observed quality of the output in terms of meaning preservation and grammaticality.

Recent applications of neural language models have led to an increased interest in the automatic generation of natural language. However impressive, the evaluation of neurally generated text has so far remained rather informal and anecdotal. Here, we present an attempt at the systematic assessment of one aspect of the quality of neurally generated text. We focus on a specific aspect of neural language generation: its ability to reproduce authorial writing styles. Using established models for authorship attribution, we empirically assess the stylistic qualities of neurally generated text. In comparison to conventional language models, neural models generate fuzzier text, that is relatively harder to attribute correctly. Nevertheless, our results also suggest that neurally generated text offers more valuable perspectives for the augmentation of training data.