We demonstrate the multilingual search engine and Ngram viewer that was built on top of the Parlamint dataset using the recently available translations. The user interface and SERP are carefully designed for querying parliamentary proceedings and for the intended use by citizens, journalists and political scholars. Demo at https://debateabase.wooverheid.nl. Keywords: Multilingual Search, Parliamentary Proceedings, Ngram Viewer, Machine Translation

The paper presents new causal commonsense reasoning datasets for South Slavic dialects, based on the Choice of Plausible Alternatives (COPA) dataset. The dialectal datasets are built by translating by native dialect speakers from the English original and the corresponding standard translation. Three dialects are covered – the Cerkno dialect of Slovenian, the Chakavian dialect of Croatian and the Torlak dialect of Serbian. The datasets are the first resource for evaluation of large language models on South Slavic dialects, as well as among the first commonsense reasoning datasets on dialects overall. The paper describes specific challenges met during the translation process. A comparison of the dialectal datasets with their standard language counterparts shows a varying level of character-level, word-level and lexicon-level deviation of dialectal text from the standard datasets. The observed differences are well reproduced in initial zero-shot and 10-shot experiments, where the Slovenian Cerkno dialect and the Croatian Chakavian dialect show significantly lower results than the Torlak dialect. These results show also for the dialectal datasets to be significantly more challenging than the standard datasets. Finally, in-context learning on just 10 examples shows to improve the results dramatically, especially for the dialects with the lowest results.

The paper presents the JSI and WüNLP systems submitted to the DIALECT-COPA shared task on causal commonsense reasoning in dialectal texts. Jointly, we compare LLM-based zero-shot and few-shot in-context inference (JSI team), and task-specific few-shot fine-tuning, in English and respective standard language, with zero-shot cross-lingual transfer (ZS-XLT) to the test dialects (WüNLP team). Given the very strong zero-shot and especially few-shot in-context learning (ICL) performance, we further investigate whether task semantics, or language/dialect semantics explain the strong performance, showing that a significant part of the improvement indeed stems from learning the language or dialect semantics from the in-context examples, with only a minor contribution from understanding the nature of the task. The higher importance of the dialect semantics to the task semantics is further shown by the finding that the in-context learning with only a few dialectal instances achieves comparable results to the supervised fine-tuning approach on hundreds of instances in standard language.

The world of language models is going through turbulent times, better and ever larger models are coming out at an unprecedented speed. However, we argue that, especially for the scientific community, encoder models of up to 1 billion parameters are still very much needed, their primary usage being in enriching large collections of data with metadata necessary for downstream research. We investigate the best way to ensure the existence of such encoder models on the set of very closely related languages - Croatian, Serbian, Bosnian and Montenegrin, by setting up a diverse benchmark for these languages, and comparing the trained-from-scratch models with the new models constructed via additional pretraining of existing multilingual models. We show that comparable performance to dedicated from-scratch models can be obtained by additionally pretraining available multilingual models even with a limited amount of computation. We also show that neighboring languages, in our case Slovenian, can be included in the additional pretraining with little to no loss in the performance of the final model.

This paper presents a collection of highly comparable web corpora of Slovenian, Croatian, Bosnian, Montenegrin, Serbian, Macedonian, and Bulgarian, covering thereby the whole spectrum of official languages in the South Slavic language space. The collection of these corpora comprises a total of 13 billion tokens of texts from 26 million documents. The comparability of the corpora is ensured by a comparable crawling setup and the usage of identical crawling and post-processing technology. All the corpora were linguistically annotated with the state-of-the-art CLASSLA-Stanza linguistic processing pipeline, and enriched with document-level genre information via the Transformer-based multilingual X-GENRE classifier, which further enhances comparability at the level of linguistic annotation and metadata enrichment. The genre-focused analysis of the resulting corpora shows a rather consistent distribution of genres throughout the seven corpora, with variations in the most prominent genre categories being well-explained by the economic strength of each language community. A comparison of the distribution of genre categories across the corpora indicates that web corpora from less developed countries primarily consist of news articles. Conversely, web corpora from economically more developed countries exhibit a smaller proportion of news content, with a greater presence of promotional and opinionated texts.

Large, curated, web-crawled corpora play a vital role in training language models (LMs). They form the lion’s share of the training data in virtually all recent LMs, such as the well-known GPT, LLaMA and XLM-RoBERTa models. However, despite this importance, relatively little attention has been given to the quality of these corpora. In this paper, we compare four of the currently most relevant large, web-crawled corpora (CC100, MaCoCu, mC4 and OSCAR) across eleven lower-resourced European languages. Our approach is two-fold: first, we perform an intrinsic evaluation by performing a human evaluation of the quality of samples taken from different corpora; then, we assess the practical impact of the qualitative differences by training specific LMs on each of the corpora and evaluating their performance on downstream tasks. We find that there are clear differences in quality of the corpora, with MaCoCu and OSCAR obtaining the best results. However, during the extrinsic evaluation, we actually find that the CC100 corpus achieves the highest scores. We conclude that, in our experiments, the quality of the web-crawled corpora does not seem to play a significant role when training LMs.

We present the most relevant results of the project MaCoCu: Massive collection and curation of monolingual and bilingual data: focus on under-resourced languages in its second year. To date, parallel and monolingual corpora have been produced for seven low-resourced European languages by crawling large amounts of textual data from selected top-level domains of the Internet; both human and automatic evaluation show its usefulness. In addition, several large language models pretrained on MaCoCu data have been published, as well as the code used to collect and curate the data.

Collecting texts from the web enables a rapid creation of monolingual and parallel corpora of unprecedented size. However, unlike manually-collected corpora, authors and end users do not know which texts make up the web collections. In this work, we analyse the content of seven European parallel web corpora, collected from national top-level domains, by analysing the English variety and genre distribution in them. We develop and provide a lexicon-based British-American variety classifier, which we use to identify the English variety. In addition, we apply a Transformer-based genre classifier to corpora to analyse genre distribution and the interplay between genres and English varieties. The results reveal significant differences among the seven corpora in terms of different genre distribution and different preference for English varieties.

Automatic discrimination between Bosnian, Croatian, Montenegrin and Serbian is a hard task due to the mutual intelligibility of these South-Slavic languages. In this paper, we introduce the BENCHić-lang benchmark for discriminating between these four languages. The benchmark consists of two datasets from different domains - a Twitter and a news dataset - selected with the aim of fostering cross-dataset evaluation of different modelling approaches. We experiment with the baseline SVM models, based on character n-grams, which perform nicely in-dataset, but do not generalize well in cross-dataset experiments. Thus, we introduce another approach, exploiting only web-crawled data and the weak supervision signal coming from the respective country/language top-level domains. The resulting simple Naive Bayes model, based on less than a thousand word features extracted from web data, outperforms the baseline models in the cross-dataset scenario and achieves good levels of generalization across datasets.

This paper presents a new training dataset for automatic genre identification GINCO, which is based on 1,125 crawled Slovenian web documents that consist of 650,000 words. Each document was manually annotated for genre with a new annotation schema that builds upon existing schemata, having primarily clarity of labels and inter-annotator agreement in mind. The dataset consists of various challenges related to web-based data, such as machine translated content, encoding errors, multiple contents presented in one document etc., enabling evaluation of classifiers in realistic conditions. The initial machine learning experiments on the dataset show that (1) pre-Transformer models are drastically less able to model the phenomena, with macro F1 metrics ranging around 0.22, while Transformer-based models achieve scores of around 0.58, and (2) multilingual Transformer models work as well on the task as the monolingual models that were previously proven to be superior to multilingual models on standard NLP tasks.

We introduce the project “MaCoCu: Massive collection and curation of monolingual and bilingual data: focus on under-resourced languages”, funded by the Connecting Europe Facility, which is aimed at building monolingual and parallel corpora for under-resourced European languages. The approach followed consists of crawling large amounts of textual data from carefully selected top-level domains of the Internet, and then applying a curation and enrichment pipeline. In addition to corpora, the project will release successive versions of the free/open-source web crawling and curation software used.