This paper presents the findings of the SIGMORPHON 2023 Shared Task on Interlinear Glossing. This first iteration of the shared task explores glossing of a set of six typologically diverse languages: Arapaho, Gitksan, Lezgi, Natügu, Tsez and Uspanteko. The shared task encompasses two tracks: a resource-scarce closed track and an open track, where participants are allowed to utilize external data resources. Five teams participated in the shared task. The winning team Tü-CL achieved a 23.99%-point improvement over a baseline RoBERTa system in the closed track and a 17.42%-point improvement in the open track.

This paper presents several different neural subword modelling based approaches to interlinear glossing for seven under-resourced languages as a part of the 2023 SIGMORPHON shared task on interlinear glossing. We experiment with various augmentation and tokenization strategies for both the open and closed tracks of data. We found that while byte-level models may perform well for greater amounts of data, character based approaches remain competitive in their performance in lower resource settings.

Data augmentation techniques are widely used in low-resource automatic morphological inflection to address the issue of data sparsity. However, the full implications of these techniques remain poorly understood. In this study, we aim to shed light on the theoretical aspects of the data augmentation strategy StemCorrupt, a method that generates synthetic examples by randomly substituting stem characters in existing gold standard training examples. Our analysis uncovers that StemCorrupt brings about fundamental changes in the underlying data distribution, revealing inherent compositional concatenative structure. To complement our theoretical analysis, we investigate the data-efficiency of StemCorrupt. Through evaluation across a diverse set of seven typologically distinct languages, we demonstrate that selecting a subset of datapoints with both high diversity and high predictive uncertainty significantly enhances the data-efficiency of compared to competitive baselines. Furthermore, we explore the impact of typological features on the choice of augmentation strategy and find that languages incorporating non-concatenativity, such as morphonological alternations, derive less benefit from synthetic examples with high predictive uncertainty. We attribute this effect to phonotactic violations induced by StemCorrupt, emphasizing the need for further research to ensure optimal performance across the entire spectrum of natural language morphology.

With a growing focus on morphological inflection systems for languages where high-quality data is scarce, training data noise is a serious but so far largely ignored concern. We aim at closing this gap by investigating the types of noise encountered within a pipeline for truly unsupervised morphological paradigm completion and its impact on morphological inflection systems: First, we propose an error taxonomy and annotation pipeline for inflection training data. Then, we compare the effect of different types of noise on multiple state-of-the- art inflection models. Finally, we propose a novel character-level masked language modeling (CMLM) pretraining objective and explore its impact on the models’ resistance to noise. Our experiments show that various architectures are impacted differently by separate types of noise, but encoder-decoders tend to be more robust to noise than models trained with a copy bias. CMLM pretraining helps transformers, but has lower impact on LSTMs.

The Universal Morphology (UniMorph) project is a collaborative effort providing broad-coverage instantiated normalized morphological inflection tables for hundreds of diverse world languages. The project comprises two major thrusts: a language-independent feature schema for rich morphological annotation, and a type-level resource of annotated data in diverse languages realizing that schema. This paper presents the expansions and improvements on several fronts that were made in the last couple of years (since McCarthy et al. (2020)). Collaborative efforts by numerous linguists have added 66 new languages, including 24 endangered languages. We have implemented several improvements to the extraction pipeline to tackle some issues, e.g., missing gender and macrons information. We have amended the schema to use a hierarchical structure that is needed for morphological phenomena like multiple-argument agreement and case stacking, while adding some missing morphological features to make the schema more inclusive. In light of the last UniMorph release, we also augmented the database with morpheme segmentation for 16 languages. Lastly, this new release makes a push towards inclusion of derivational morphology in UniMorph by enriching the data and annotation schema with instances representing derivational processes from MorphyNet.

This paper presents a new inflectional resource for Gitksan, a low-resource Indigenous language of Canada. We use Gitksan data in interlinear glossed format, stemming from language documentation efforts, to build a database of partial inflection tables. We then enrich this morphological resource by filling in blank slots in the partial inflection tables using neural transformer reinflection models. We extend the training data for our transformer reinflection models using two data augmentation techniques: data hallucination and back-translation. Experimental results demonstrate substantial improvements from data augmentation, with data hallucination delivering particularly impressive gains. We also release reinflection models for Gitksan.

Data augmentation strategies are increasingly important in NLP pipelines for low-resourced and endangered languages, and in neural morphological inflection, augmentation by so called data hallucination is a popular technique. This paper presents a detailed analysis of inflection models trained with and without data hallucination for the low-resourced Canadian Indigenous language Gitksan. Our analysis reveals evidence for a concatenative inductive bias in augmented models—in contrast to models trained without hallucination, they strongly prefer affixing inflection patterns over suppletive ones. We find that preference for affixation in general improves inflection performance in “wug test” like settings, where the model is asked to inflect lexemes missing from the training set. However, data hallucination dramatically reduces prediction accuracy for reduplicative forms due to a misanalysis of reduplication as affixation. While the overall impact of data hallucination for unseen lexemes remains positive, our findings call for greater qualitative analysis and more varied evaluation conditions in testing automatic inflection systems. Our results indicate that further innovations in data augmentation for computational morphology are desirable.

This paper explores a special case in multilingual machine translation: so called multi-parallel translation, where the target data for all language pairs are identical. While multi-parallelism offers benefits which are not available in a standard translation setting, translation models can easily overfit when training data are limited. We introduce a regularizer, the divergence penalty, which penalizes the translation model when it represents source sentences with identical target translations in divergent ways. Experiments on very low-resourced Indigenous North American languages show that an initially deficient multilingual translator can improve by 4.9 BLEU through mBART pre-training, and 5.5 BLEU points with the strategic addition of monolingual data, and that a divergence penalty leads to further increases of 0.4 BLEU. Further experiments on Germanic languages demonstrate a improvement of 0.5 BLEU when applying the divergence penalty. An investigation of the neural encoder representations learned by our translation models shows that the divergence penalty encourages models to learn a unified neural interlingua.

Translating between languages where certain features are marked morphologically in one but absent or marked contextually in the other is an important test case for machine translation. When translating into English which marks (in)definiteness morphologically, from Yorùbá which uses bare nouns but marks these features contextually, ambiguities arise. In this work, we perform fine-grained analysis on how an SMT system compares with two NMT systems (BiLSTM and Transformer) when translating bare nouns in Yorùbá into English. We investigate how the systems what extent they identify BNs, correctly translate them, and compare with human translation patterns. We also analyze the type of errors each model makes and provide a linguistic description of these errors. We glean insights for evaluating model performance in low-resource settings. In translating bare nouns, our results show the transformer model outperforms the SMT and BiLSTM models for 4 categories, the BiLSTM outperforms the SMT model for 3 categories while the SMT outperforms the NMT models for 1 category.

In this study, we propose a morpheme-based scheme for Korean dependency parsing and adopt the proposed scheme to Universal Dependencies. We present the linguistic rationale that illustrates the motivation and the necessity of adopting the morpheme-based format, and develop scripts that convert between the original format used by Universal Dependencies and the proposed morpheme-based format automatically. The effectiveness of the proposed format for Korean dependency parsing is then testified by both statistical and neural models, including UDPipe and Stanza, with our carefully constructed morpheme-based word embedding for Korean. morphUD outperforms parsing results for all Korean UD treebanks, and we also present detailed error analysis.

This paper presents experiments on morphological inflection using data from the SIGMORPHON-UniMorph 2022 Shared Task 0: Generalization and Typologically Diverse Morphological Inflection. We present a transformer inflection system, which enriches the standard transformer architecture with reverse positional encoding and type embeddings. We further apply data hallucination and lemma copying to augment training data. We train models using a two-stage procedure: (1) We first train on the augmented training data using standard backpropagation and teacher forcing. (2) We then continue training with a variant of the scheduled sampling algorithm dubbed student forcing. Our system delivers competitive performance under the small and large data conditions on the shared task datasets.

Automatic morphological processing can aid downstream natural language processing applications, especially for low-resource languages, and assist language documentation efforts for endangered languages. Having long been multilingual, the field of computational morphology is increasingly moving towards approaches suitable for languages with minimal or no annotated resources. First, we survey recent developments in computational morphology with a focus on low-resource languages. Second, we argue that the field is ready to tackle the logical next challenge: understanding a language’s morphology from raw text alone. We perform an empirical study on a truly unsupervised version of the paradigm completion task and show that, while existing state-of-the-art models bridged by two newly proposed models we devise perform reasonably, there is still much room for improvement. The stakes are high: solving this task will increase the language coverage of morphological resources by a number of magnitudes.

Recent progress in NLP is driven by pretrained models leveraging massive datasets and has predominantly benefited the world’s political and economic superpowers. Technologically underserved languages are left behind because they lack such resources. Hundreds of underserved languages, nevertheless, have available data sources in the form of interlinear glossed text (IGT) from language documentation efforts. IGT remains underutilized in NLP work, perhaps because its annotations are only semi-structured and often language-specific. With this paper, we make the case that IGT data can be leveraged successfully provided that target language expertise is available. We specifically advocate for collaboration with documentary linguists. Our paper provides a roadmap for successful projects utilizing IGT data: (1) It is essential to define which NLP tasks can be accomplished with the given IGT data and how these will benefit the speech community. (2) Great care and target language expertise is required when converting the data into structured formats commonly employed in NLP. (3) Task-specific and user-specific evaluation can help to ascertain that the tools which are created benefit the target language speech community. We illustrate each step through a case study on developing a morphological reinflection system for the Tsimchianic language Gitksan.

We present the University of British Columbia’s submission to the MRL shared task on multilingual clause-level morphology. Our submission extends word-level inflectional models to the clause-level in two ways: first, by evaluating the role that BPE has on the learning of inflectional morphology, and second, by evaluating the importance of a copy bias obtained through data hallucination. Experiments demonstrate a strong preference for language-tuned BPE and a copy bias over a vanilla transformer. The methods are complementary for inflection and analysis tasks – combined models see error reductions of 38% for inflection and 15.6% for analysis; However, this synergy does not hold for reinflection, which performs best under a BPE-only setting. A deeper analysis of the errors generated by our models illustrates that the copy bias may be too strong - the combined model produces predictions more similar to the copy-influenced system, despite the success of the BPE-model.

Sequence-to-sequence models have delivered impressive results in word formation tasks such as morphological inflection, often learning to model subtle morphophonological details with limited training data. Despite the performance, the opacity of neural models makes it difficult to determine whether complex generalizations are learned, or whether a kind of separate rote memorization of each morphophonological process takes place. To investigate whether complex alternations are simply memorized or whether there is some level of generalization across related sound changes in a sequence-to-sequence model, we perform several experiments on Finnish consonant gradation—a complex set of sound changes triggered in some words by certain suffixes. We find that our models often—though not always—encode 17 different consonant gradation processes in a handful of dimensions in the RNN. We also show that by scaling the activations in these dimensions we can control whether consonant gradation occurs and the direction of the gradation.

We describe the second SIGMORPHON shared task on unsupervised morphology: the goal of the SIGMORPHON 2021 Shared Task on Unsupervised Morphological Paradigm Clustering is to cluster word types from a raw text corpus into paradigms. To this end, we release corpora for 5 development and 9 test languages, as well as gold partial paradigms for evaluation. We receive 14 submissions from 4 teams that follow different strategies, and the best performing system is based on adaptor grammars. Results vary significantly across languages. However, all systems are outperformed by a supervised lemmatizer, implying that there is still room for improvement.

This paper describes the submission of the CU-UBC team for the SIGMORPHON 2021 Shared Task 2: Unsupervised morphological paradigm clustering. Our system generates paradigms using morphological transformation rules which are discovered from raw data. We experiment with two methods for discovering rules. Our first approach generates prefix and suffix transformations between similar strings. Secondly, we experiment with more general rules which can apply transformations inside the input strings in addition to prefix and suffix transformations. We find that the best overall performance is delivered by prefix and suffix rules but more general transformation rules perform better for languages with templatic morphology and very high morpheme-to-word ratios.

This paper presents a finite-state morphological analyzer for the Gitksan language. The analyzer draws from a 1250-token Eastern dialect wordlist. It is based on finite-state technology and additionally includes two extensions which can provide analyses for out-of-vocabulary words: rules for generating predictable dialect variants, and a neural guesser component. The pre-neural analyzer, tested against interlinear-annotated texts from multiple dialects, achieves coverage of (75-81%), and maintains high precision (95-100%). The neural extension improves coverage at the cost of lowered precision.

Akkadian was an East-Semitic language spoken in ancient Mesopotamia. The language is attested on hundreds of thousands of cuneiform clay tablets. Several Akkadian text corpora contain only the transliterated text. In this paper, we investigate automated phonological transcription of the transliterated corpora. The phonological transcription provides a linguistically appealing form to represent Akkadian, because the transcription is normalized according to the grammatical description of a given dialect and explicitly shows the Akkadian renderings for Sumerian logograms. Because cuneiform text does not mark the inflection for logograms, the inflected form needs to be inferred from the sentence context. To the best of our knowledge, this is the first documented attempt to automatically transcribe Akkadian. Using a context-aware neural network model, we are able to automatically transcribe syllabic tokens at near human performance with 96% recall @ 3, while the logogram transcription remains more challenging at 82% recall @ 3.

Akkadian is a fairly well resourced extinct language that does not yet have a comprehensive morphological analyzer available. In this paper we describe a general finite-state based morphological model for Babylonian, a southern dialect of the Akkadian language, that can achieve a coverage up to 97.3% and recall up to 93.7% on lemmatization and POS-tagging task on token level from a transcribed input. Since Akkadian word forms exhibit a high degree of morphological ambiguity, in that only 20.1% of running word tokens receive a single unambiguous analysis, we attempt a first pass at weighting our finite-state transducer, using existing extensive Akkadian corpora which have been partially validated for their lemmas and parts-of-speech but not the entire morphological analyses. The resultant weighted finite-state transducer yields a moderate improvement so that for 57.4% of the word tokens the highest ranked analysis is the correct one. We conclude with a short discussion on how morphological ambiguity in the analysis of Akkadian could be further reduced with improvements in the training data used in weighting the finite-state transducer as well as through other, context-based techniques.

The Universal Morphology (UniMorph) project is a collaborative effort providing broad-coverage instantiated normalized morphological paradigms for hundreds of diverse world languages. The project comprises two major thrusts: a language-independent feature schema for rich morphological annotation and a type-level resource of annotated data in diverse languages realizing that schema. We have implemented several improvements to the extraction pipeline which creates most of our data, so that it is both more complete and more correct. We have added 66 new languages, as well as new parts of speech for 12 languages. We have also amended the schema in several ways. Finally, we present three new community tools: two to validate data for resource creators, and one to make morphological data available from the command line. UniMorph is based at the Center for Language and Speech Processing (CLSP) at Johns Hopkins University in Baltimore, Maryland. This paper details advances made to the schema, tooling, and dissemination of project resources since the UniMorph 2.0 release described at LREC 2018.

A broad goal in natural language processing (NLP) is to develop a system that has the capacity to process any natural language. Most systems, however, are developed using data from just one language such as English. The SIGMORPHON 2020 shared task on morphological reinflection aims to investigate systems’ ability to generalize across typologically distinct languages, many of which are low resource. Systems were developed using data from 45 languages and just 5 language families, fine-tuned with data from an additional 45 languages and 10 language families (13 in total), and evaluated on all 90 languages. A total of 22 systems (19 neural) from 10 teams were submitted to the task. All four winning systems were neural (two monolingual transformers and two massively multilingual RNN-based models with gated attention). Most teams demonstrate utility of data hallucination and augmentation, ensembles, and multilingual training for low-resource languages. Non-neural learners and manually designed grammars showed competitive and even superior performance on some languages (such as Ingrian, Tajik, Tagalog, Zarma, Lingala), especially with very limited data. Some language families (Afro-Asiatic, Niger-Congo, Turkic) were relatively easy for most systems and achieved over 90% mean accuracy while others were more challenging.

The task of grapheme-to-phoneme (G2P) conversion is important for both speech recognition and synthesis. Similar to other speech and language processing tasks, in a scenario where only small-sized training data are available, learning G2P models is challenging. We describe a simple approach of exploiting model ensembles, based on multilingual Transformers and self-training, to develop a highly effective G2P solution for 15 languages. Our models are developed as part of our participation in the SIGMORPHON 2020 Shared Task 1 focused at G2P. Our best models achieve 14.99 word error rate (WER) and 3.30 phoneme error rate (PER), a sizeable improvement over the shared task competitive baselines.

Morphological inflection, like many sequence-to-sequence tasks, sees great performance from recurrent neural architectures when data is plentiful, but performance falls off sharply in lower-data settings. We investigate one aspect of neural seq2seq models that we hypothesize contributes to overfitting - teacher forcing. By creating different training and test conditions, exposure bias increases the likelihood that a system too closely models its training data. Experiments show that teacher-forced models struggle to recover when they enter unknown territory. However, a simple modification to the training algorithm to more closely mimic test conditions creates models that are better able to generalize to unseen environments.

In this paper, we present the findings of the Third VarDial Evaluation Campaign organized as part of the sixth edition of the workshop on Natural Language Processing (NLP) for Similar Languages, Varieties and Dialects (VarDial), co-located with NAACL 2019. This year, the campaign included five shared tasks, including one task re-run – German Dialect Identification (GDI) – and four new tasks – Cross-lingual Morphological Analysis (CMA), Discriminating between Mainland and Taiwan variation of Mandarin Chinese (DMT), Moldavian vs. Romanian Cross-dialect Topic identification (MRC), and Cuneiform Language Identification (CLI). A total of 22 teams submitted runs across the five shared tasks. After the end of the competition, we received 14 system description papers, which are published in the VarDial workshop proceedings and referred to in this report.

The SIGMORPHON 2019 shared task on cross-lingual transfer and contextual analysis in morphology examined transfer learning of inflection between 100 language pairs, as well as contextual lemmatization and morphosyntactic description in 66 languages. The first task evolves past years’ inflection tasks by examining transfer of morphological inflection knowledge from a high-resource language to a low-resource language. This year also presents a new second challenge on lemmatization and morphological feature analysis in context. All submissions featured a neural component and built on either this year’s strong baselines or highly ranked systems from previous years’ shared tasks. Every participating team improved in accuracy over the baselines for the inflection task (though not Levenshtein distance), and every team in the contextual analysis task improved on both state-of-the-art neural and non-neural baselines.

We investigate different ensemble learning techniques for neural morphological inflection using bidirectional LSTM encoder-decoder models with attention. We experiment with weighted and unweighted majority voting and bagging. We find that all investigated ensemble methods lead to improved accuracy over a baseline of a single model. However, contrary to expectation based on earlier work by Najafi et al. (2018) and Silfverberg et al. (2017), weighting does not deliver clear benefits. Bagging was found to underperform plain voting ensembles in general.

We conduct a manual error analysis of the CoNLL-SIGMORPHON Shared Task on Morphological Reinflection. This task involves natural language generation: systems are given a word in citation form (e.g., hug) and asked to produce the corresponding inflected form (e.g., the simple past hugged). We propose an error taxonomy and use it to annotate errors made by the top two systems across twelve languages. Many of the observed errors are related to inflectional patterns sensitive to inherent linguistic properties such as animacy or affect; many others are failures to predict truly unpredictable inflectional behaviors. We also find nearly one quarter of the residual “errors” reflect errors in the gold data.

This paper presents the submission of the UH&CU team (Joint University of Colorado and University of Helsinki team) for the VarDial 2018 shared task on morphosyntactic tagging of Croatian, Slovenian and Serbian tweets. Our system is a bidirectional LSTM tagger which emits tags as character sequences using an LSTM generator in order to be able to handle unknown tags and combinations of several tags for one token which occur in the shared task data sets. To the best of our knowledge, using an LSTM generator is a novel approach. The system delivers sizable improvements of more than 6%-points over a baseline trigram tagger. Overall, the performance of our system is quite even for all three languages.

Modeling morphological inflection is an important task in Natural Language Processing. In contrast to earlier work that has largely used orthographic representations, we experiment with this task in a phonetic character space, representing inputs as either IPA segments or bundles of phonological distinctive features. We show that both of these inputs, somewhat counterintuitively, achieve similar accuracies on morphological inflection, slightly lower than orthographic models. We conclude that providing detailed phonological representations is largely redundant when compared to IPA segments, and that articulatory distinctions relevant for word inflection are already latently present in the distributional properties of many graphemic writing systems.

The Universal Dependencies (UD) and Universal Morphology (UniMorph) projects each present schemata for annotating the morphosyntactic details of language. Each project also provides corpora of annotated text in many languages—UD at the token level and UniMorph at the type level. As each corpus is built by different annotators, language-specific decisions hinder the goal of universal schemata. With compatibility of tags, each project’s annotations could be used to validate the other’s. Additionally, the availability of both type- and token-level resources would be a boon to tasks such as parsing and homograph disambiguation. To ease this interoperability, we present a deterministic mapping from Universal Dependencies v2 features into the UniMorph schema. We validate our approach by lookup in the UniMorph corpora and find a macro-average of 64.13% recall. We also note incompatibilities due to paucity of data on either side. Finally, we present a critical evaluation of the foundations, strengths, and weaknesses of the two annotation projects.

The Paradigm Cell Filling Problem in morphology asks to complete word inflection tables from partial ones. We implement novel neural models for this task, evaluating them on 18 data sets in 8 languages, showing performance that is comparable with previous work with far less training data. We also publish a new dataset for this task and code implementing the system described in this paper.

In supervised learning of morphological patterns, the strategy of generalizing inflectional tables into more abstract paradigms through alignment of the longest common subsequence found in an inflection table has been proposed as an efficient method to deduce the inflectional behavior of unseen word forms. In this paper, we extend this notion of morphological ‘paradigm’ from earlier work and provide a formalization that more accurately matches linguist intuitions about what an inflectional paradigm is. Additionally, we propose and evaluate a mechanism for learning full human-readable paradigm specifications from incomplete data—a scenario when we only have access to a few inflected forms for each lexeme, and want to reconstruct the missing inflections as well as generalize and group the witnessed patterns into a model of more abstract paradigmatic behavior of lexemes.

Most NLP resources that offer annotations at the word segment level provide morphological annotation that includes features indicating tense, aspect, modality, gender, case, and other inflectional information. Such information is rarely aligned to the relevant parts of the words—i.e. the allomorphs, as such annotation would be very costly. These unaligned weak labelings are commonly provided by annotated NLP corpora such as treebanks in various languages. Although they lack alignment information, the presence/absence of labels at the word level is also consistent with the amount of supervision assumed to be provided to L1 and L2 learners. In this paper, we explore several methods to learn this latent alignment between parts of word forms and the grammatical information provided. All the methods under investigation favor hypotheses regarding allomorphs of morphemes that re-use a small inventory, i.e. implicitly minimize the number of allomorphs that a morpheme can be realized as. We show that the provided information offers a significant advantage for both word segmentation and the learning of allomorphy.

Flag diacritics, which are special multi-character symbols executed at runtime, enable optimising finite-state networks by combining identical sub-graphs of its transition graph. Traditionally, the feature has required linguists to devise the optimisations to the graph by hand alongside the morphological description. In this paper, we present a novel method for discovering flag positions in morphological lexicons automatically, based on the morpheme structure implicit in the language description. With this approach, we have gained significant decrease in the size of finite-state networks while maintaining reasonable application speed. The algorithm can be applied to any language description, where the biggest achievements are expected in large and complex morphologies. The most noticeable reduction in size we got with a morphological transducer for Greenlandic, whose original size is on average about 15 times larger than other morphologies. With the presented hyper-minimization method, the transducer is reduced to 10,1% of the original size, with lookup speed decreased only by 9,5%.