In this study we approach the detection of null subjects and impersonal constructions in Spanish using a machine translation methodology. We repurpose the Spanish AnCora corpus, converting it to a parallel set that transforms Spanish sentences into a format that allows us to detect and classify verbs, and train LSTM-based neural machine translation systems to perform this task. Various models differing on output format and hyperparameters were evaluated. Experimental results proved this approach to be highly resource-effective, obtaining results comparable to or surpassing the state of the art found in existing literature, while employing modest computational resources. Additionally, an improved dataset for training and evaluating Spanish null-subject detection tools was elaborated for this project, that could aid in the creation and serve as a benchmark for further developments in the area.

This paper presents the results of the first shared task about the creation of educational materials for three indigenous languages of the Americas.The task proposes to automatically generate variations of sentences according to linguistic features that could be used for grammar exercises.The languages involved in this task are Bribri, Maya, and Guarani.Seven teams took part in the challenge, submitting a total of 22 systems, obtaining very promising results.

This paper presents the findings of the third iteration of the AmericasNLP Shared Task on Machine Translation. This year’s competition features eleven Indigenous languages found across North, Central, and South America. A total of six teams participate with a total of 157 submissions across all languages and models. Two baselines – the Sheffield and Helsinki systems from 2023 – are provided and represent hard-to-beat starting points for the competition. In addition to the baselines, teams are given access to a new repository of training data which consists of data collected by teams in prior shared tasks. Using ChrF++ as the main competition metric, we see improvements over the baseline for 4 languages: Chatino, Guarani, Quechua, and Rarámuri, with performance increases over the best baseline of 4.2 ChrF++. In this work, we present a summary of the submitted systems, results, and a human evaluation of system outputs for Bribri, which consists of both (1) a rating of meaning and fluency and (2) a qualitative error analysis of outputs from the best submitted system.

One of the main problems low-resource languages face in NLP can be pictured as a vicious circle: data is needed to build and test tools, but the available text is scarce and there are not powerful tools to collect it.In order to break this circle for Guarani, we explore if text automatically generated from a grammar can work as a Data Augmentation technique to boost the performance of Guarani-Spanish Machine Translation (MT) systems.After building a grammar-based system that generates Spanish text and syntactically transfers it to Guarani, we perform several experiments by pretraining models using this synthetic text.We find that the MT systems that are pretrained with synthetic text perform better, even outperforming previous baselines.

This paper presents a language model trained from scratch exclusively on a brand new corpus consisting of about 6 GiB of Uruguayan newspaper text. We trained the model for 30 days on a single Nvidia P100 using the RoBERTa-base architecture but with considerably fewer parameters than other standard RoBERTa models. We evaluated the model on two NLP tasks and found that it outperforms BETO, the widely used Spanish BERT pre-trained model. We also compared our model on the masked-word prediction task with two popular multilingual BERT-based models, Multilingual BERT and XLM-RoBERTa, obtaining outstanding results on sentences from the Uruguayan press domain. Our experiments show that training a language model on a domain-specific corpus can significantly improve performance even when the model is smaller and was trained with significantly less data than more standard pre-trained models.

Crosswords are a powerful tool that could be used in educational contexts, but they are not that easy to build. In this work, we present experiments on automatically extracting clues from simple texts that could be used to create crosswords, with the aim of using them in the context of teaching English at the beginner level. We present a series of heuristic patterns based on NLP tools for extracting clues, and use them to create a set of 2209 clues from a collection of 400 simple texts. Human annotators labeled the clues, and this dataset is used to evaluate the performance of our heuristics, and also to create a classifier that predicts if an extracted clue is correct. Our best classifier achieves an accuracy of 84%.

In this paper we present the participation of the RETUYT-INCO team at the BEA-MLSP 2024 shared task. We followed different approaches, from Multilayer Perceptron models with word embeddings to Large Language Models fine-tuned on different datasets: already existing, crowd-annotated, and synthetic.Our best models are based on fine-tuning Mistral-7B, either with a manually annotated dataset or with synthetic data.

This paper presents a work in progress about creating a Guarani version of the WordNet database. Guarani is an indigenous South American language and is a low-resource language from the NLP perspective. Following the expand approach, we aim to find Guarani lemmas that correspond to the concepts defined in WordNet. We do this through three strategies that try to select the correct lemmas from Guarani-Spanish datasets. We ran them through three different bilingual dictionaries and had native speakers assess the results. This procedure found Guarani lemmas for about 6.5 thousand synsets, including 27% of the base WordNet concepts. However, more work on the quality of the selected words will be needed in order to create a final version of the dataset.

This paper presents the results of our participation in the BEA 2023 shared task, which focuses on generating AI teacher responses in educational dialogues. We conducted experiments using several Open-Source Large Language Models (LLMs) and explored fine-tuning techniques along with prompting strategies, including Few-Shot and Chain-of-Thought approaches. Our best model was ranked 4.5 in the competition with a BertScore F1 of 0.71 and a DialogRPT final (avg) of 0.35. Nevertheless, our internal results did not exactly correlate with those obtained in the competition, which showed the difficulty in evaluating this task. Other challenges we faced were data leakage on the train set and the irregular format of the conversations.

Large multilingual models have inspired a new class of word alignment methods, which work well for the model’s pretraining languages. However, the languages most in need of automatic alignment are low-resource and, thus, not typically included in the pretraining data. In this work, we ask: How do modern aligners perform on unseen languages, and are they better than traditional methods? We contribute gold-standard alignments for Bribri–Spanish, Guarani–Spanish, Quechua–Spanish, and Shipibo-Konibo–Spanish. With these, we evaluate state-of-the-art aligners with and without model adaptation to the target language. Finally, we also evaluate the resulting alignments extrinsically through two downstream tasks: named entity recognition and part-of-speech tagging. We find that although transformer-based methods generally outperform traditional models, the two classes of approach remain competitive with each other.

This paper examines the use of manually part-of-speech tagged sign language gloss data in the Text2Gloss and Gloss2Text translation tasks, as well as running an LSTM-based sequence labelling model on the same glosses for automatic part-of-speech tagging. We find that a combination of tag-enhanced glosses and pretraining the neural model positively impacts performance in the translation tasks. The results of the tagging task are limited, but provide a methodological framework for further research into tagging sign language gloss data.

This paper presents a series of experiments on translating between spoken Spanish and Spanish Sign Language glosses (LSE), including enriching Neural Machine Translation (NMT) systems with linguistic features, and creating synthetic data to pretrain and later on finetune a neural translation model. We found evidence that pretraining over a large corpus of LSE synthetic data aligned to Spanish sentences could markedly improve the performance of the translation models.

This work presents a parallel corpus of Guarani-Spanish text aligned at sentence level. The corpus contains about 30,000 sentence pairs, and is structured as a collection of subsets from different sources, further split into training, development and test sets. A sample of sentences from the test set was manually annotated by native speakers in order to incorporate meta-linguistic annotations about the Guarani dialects present in the corpus and also the correctness of the alignment and translation. We also present some baseline MT experiments and analyze the results in terms of the subsets. We hope this corpus can be used as a benchmark for testing Guarani-Spanish MT systems, and aim to expand and improve the quality of the corpus in future iterations.

Pretrained multilingual models are able to perform cross-lingual transfer in a zero-shot setting, even for languages unseen during pretraining. However, prior work evaluating performance on unseen languages has largely been limited to low-level, syntactic tasks, and it remains unclear if zero-shot learning of high-level, semantic tasks is possible for unseen languages. To explore this question, we present AmericasNLI, an extension of XNLI (Conneau et al., 2018) to 10 Indigenous languages of the Americas. We conduct experiments with XLM-R, testing multiple zero-shot and translation-based approaches. Additionally, we explore model adaptation via continued pretraining and provide an analysis of the dataset by considering hypothesis-only models. We find that XLM-R’s zero-shot performance is poor for all 10 languages, with an average performance of 38.48%. Continued pretraining offers improvements, with an average accuracy of 43.85%. Surprisingly, training on poorly translated data by far outperforms all other methods with an accuracy of 49.12%.

We present a web application for creating games and exercises for teaching English as a foreign language with the help of NLP tools. The application contains different kinds of games such as crosswords, word searches, a memory game, and a multiplayer game based on the classic battleship pen and paper game. This application was built with the aim of supporting teachers in rural schools that are teaching English lessons, so they can easily create interactive and engaging activities for their students. We present the context and history of the project, the current state of the web application, and some ideas on how we will expand it in the future.

Machine translation for low-resource languages, such as Guarani, is a challenging task due to the lack of data. One way of tackling it is using pretrained word embeddings for model initialization. In this work we try to check if currently available data is enough to train rich embeddings for enhancing MT for Guarani and Spanish, by building a set of word embedding collections and training MT systems using them. We found that the trained vectors are strong enough to slightly improve the performance of some of the translation models and also to speed up the training convergence.

SemEval 2021 Task 7, HaHackathon, was the first shared task to combine the previously separate domains of humor detection and offense detection. We collected 10,000 texts from Twitter and the Kaggle Short Jokes dataset, and had each annotated for humor and offense by 20 annotators aged 18-70. Our subtasks were binary humor detection, prediction of humor and offense ratings, and a novel controversy task: to predict if the variance in the humor ratings was higher than a specific threshold. The subtasks attracted 36-58 submissions, with most of the participants choosing to use pre-trained language models. Many of the highest performing teams also implemented additional optimization techniques, including task-adaptive training and adversarial training. The results suggest that the participating systems are well suited to humor detection, but that humor controversy is a more challenging task. We discuss which models excel in this task, which auxiliary techniques boost their performance, and analyze the errors which were not captured by the best systems.

While Guarani is widely spoken in South America, obtaining a large amount of Guarani text from the web is hard. We present the building process of a Guarani corpus composed of a parallel Guarani-Spanish set of news articles, and a monolingual set of tweets. We perform some word embeddings experiments aiming at evaluating the quality of the Guarani split of the corpus, finding encouraging results but noticing that more diversity in text domains might be needed for further improvements.

This paper presents the results of the 2021 Shared Task on Open Machine Translation for Indigenous Languages of the Americas. The shared task featured two independent tracks, and participants submitted machine translation systems for up to 10 indigenous languages. Overall, 8 teams participated with a total of 214 submissions. We provided training sets consisting of data collected from various sources, as well as manually translated sentences for the development and test sets. An official baseline trained on this data was also provided. Team submissions featured a variety of architectures, including both statistical and neural models, and for the majority of languages, many teams were able to considerably improve over the baseline. The best performing systems achieved 12.97 ChrF higher than baseline, when averaged across languages.

This paper presents the development of a deep parser for Spanish that uses a HPSG grammar and returns trees that contain both syntactic and semantic information. The parsing process uses a top-down approach implemented using LSTM neural networks, and achieves good performance results in terms of syntactic constituency and dependency metrics, and also SRL. We describe the grammar, corpus and implementation of the parser. Our process outperforms a CKY baseline and other Spanish parsers in terms of global metrics and also for some specific Spanish phenomena, such as clitics reduplication and relative referents.

This paper presents the development of a Guarani - Spanish parallel corpus with sentence-level alignment. The Guarani sentences of the corpus use the Jopara Guarani dialect, the dialect of Guarani spoken in Paraguay, which is based on Guarani grammar and may include several Spanish loanwords or neologisms. The corpus has around 14,500 sentence pairs aligned using a semi-automatic process, containing 228,000 Guarani tokens and 336,000 Spanish tokens extracted from web sources.

This paper presents the development of a corpus of 30,000 Spanish tweets that were crowd-annotated with humor value and funniness score. The corpus contains approximately 38.6% of humorous tweets with an average score of 2.04 in a scale from 1 to 5 for the humorous tweets. The corpus has been used in an automatic humor recognition and analysis competition, obtaining encouraging results from the participants.

Related work sections or literature reviews are an essential part of every scientific article being crucial for paper reviewing and assessment. The automatic generation of related work sections can be considered an instance of the multi-document summarization problem. In order to allow the study of this specific problem, we have developed a manually annotated, machine readable data-set of related work sections, cited papers (e.g. references) and sentences, together with an additional layer of papers citing the references. We additionally present experiments on the identification of cited sentences, using as input citation contexts. The corpus alongside the gold standard are made available for use by the scientific community.

Computational Humor involves several tasks, such as humor recognition, humor generation, and humor scoring, for which it is useful to have human-curated data. In this work we present a corpus of 27,000 tweets written in Spanish and crowd-annotated by their humor value and funniness score, with about four annotations per tweet, tagged by 1,300 people over the Internet. It is equally divided between tweets coming from humorous and non-humorous accounts. The inter-annotator agreement Krippendorff’s alpha value is 0.5710. The dataset is available for general usage and can serve as a basis for humor detection and as a first step to tackle subjectivity.

We present some strategies for improving the Spanish version of WordNet, part of the MCR, selecting new lemmas for the Spanish synsets by translating the lemmas of the corresponding English synsets. We used four simple selectors that resulted in a considerable improvement of the Spanish WordNet coverage, but with relatively lower precision, then we defined two context based selectors that improved the precision of the translations.

In the current context of scientific information overload, text mining tools are of paramount importance for researchers who have to read scientific papers and assess their value. Current citation networks, which link papers by citation relationships (reference and citing paper), are useful to quantitatively understand the value of a piece of scientific work, however they are limited in that they do not provide information about what specific part of the reference paper the citing paper is referring to. This qualitative information is very important, for example, in the context of current community-based scientific summarization activities. In this paper, and relying on an annotated dataset of co-citation sentences, we carry out a number of experiments aimed at, given a citation sentence, automatically identify a part of a reference paper being cited. Additionally our algorithm predicts the specific reason why such reference sentence has been cited out of five possible reasons.

Although there are currently several versions of Princeton WordNet for different languages, the lack of development of some of these versions does not make it possible to use them in different Natural Language Processing applications. So is the case of the Spanish Wordnet contained in the Multilingual Central Repository (MCR), which we tried unsuccessfully to incorporate into an anaphora resolution application and also in search terms expansion. In this situation, different strategies to improve MCR Spanish WordNet coverage were put forward and tested, obtaining encouraging results. A specific process was conducted to increase the number of adverbs, and a few simple processes were applied which made it possible to increase, at a very low cost, the number of terms in the Spanish WordNet. Finally, a more complex method based on distributional semantics was proposed, using the relations between English Wordnet synsets, also returning positive results.