African languages have far less in-language content available digitally, making it challenging for question answering systems to satisfy the information needs of users. Cross-lingual open-retrieval question answering (XOR QA) systems – those that retrieve answer content from other languages while serving people in their native language—offer a means of filling this gap. To this end, we create Our Dataset, the first cross-lingual QA dataset with a focus on African languages. Our Dataset includes 12,000+ XOR QA examples across 10 African languages. While previous datasets have focused primarily on languages where cross-lingual QA augments coverage from the target language, Our Dataset focuses on languages where cross-lingual answer content is the only high-coverage source of answer content. Because of this, we argue that African languages are one of the most important and realistic use cases for XOR QA. Our experiments demonstrate the poor performance of automatic translation and multilingual retrieval methods. Overall, Our Dataset proves challenging for state-of-the-art QA models. We hope that the dataset enables the development of more equitable QA technology.

Welsh is an official language in Wales and is spoken by an estimated 884,300 people (29.2% of the population of Wales). Despite this status and estimated increase in speaker numbers since the last (2011) census, Welsh remains a minority language undergoing revitalisation and promotion by Welsh Government and relevant stakeholders. As part of the effort to increase the availability of Welsh digital technology, this paper introduces the first Welsh summarisation dataset, which we provide freely for research purposes to help advance the work on Welsh summarisation. The dataset was created by Welsh speakers through manually summarising Welsh Wikipedia articles. In addition, the paper discusses the implementation and evaluation of different summarisation systems for Welsh. The summarisation systems and results will serve as benchmarks for the development of summarisers in other minority language contexts.

This work presents a standard Igbo named entity recognition (IgboNER) dataset as well as the results from training and fine-tuning state-of-the-art transformer IgboNER models. We discuss the process of our dataset creation - data collection and annotation and quality checking. We also present experimental processes involved in building an IgboBERT language model from scratch as well as fine-tuning it along with other non-Igbo pre-trained models for the downstream IgboNER task. Our results show that, although the IgboNER task benefited hugely from fine-tuning large transformer model, fine-tuning a transformer model built from scratch with comparatively little Igbo text data seems to yield quite decent results for the IgboNER task. This work will contribute immensely to IgboNLP in particular as well as the wider African and low-resource NLP efforts Keywords: Igbo, named entity recognition, BERT models, under-resourced, dataset

As part of the effort to increase the availability of Welsh digital technology, this paper introduces the first human vs metrics Welsh summarisation evaluation results and dataset, which we provide freely for research purposes to help advance the work on Welsh summarisation. The system summaries were created using an extractive graph-based Welsh summariser. The system summaries were evaluated by both human and a range of ROUGE metric variants (e.g. ROUGE 1, 2, L and SU4). The summaries and evaluation results will serve as benchmarks for the development of summarisers and evaluation metrics in other minority language contexts.

African languages are spoken by over a billion people, but they are under-represented in NLP research and development. Multiple challenges exist, including the limited availability of annotated training and evaluation datasets as well as the lack of understanding of which settings, languages, and recently proposed methods like cross-lingual transfer will be effective. In this paper, we aim to move towards solutions for these challenges, focusing on the task of named entity recognition (NER). We present the creation of the largest to-date human-annotated NER dataset for 20 African languages. We study the behaviour of state-of-the-art cross-lingual transfer methods in an Africa-centric setting, empirically demonstrating that the choice of source transfer language significantly affects performance. While much previous work defaults to using English as the source language, our results show that choosing the best transfer language improves zero-shot F1 scores by an average of 14% over 20 languages as compared to using English.

We take a step towards addressing the under- representation of the African continent in NLP research by bringing together different stakeholders to create the first large, publicly available, high-quality dataset for named entity recognition (NER) in ten African languages. We detail the characteristics of these languages to help researchers and practitioners better understand the challenges they pose for NER tasks. We analyze our datasets and conduct an extensive empirical evaluation of state- of-the-art methods across both supervised and transfer learning settings. Finally, we release the data, code, and models to inspire future research on African NLP.1

Research in NLP lacks geographic diversity, and the question of how NLP can be scaled to low-resourced languages has not yet been adequately solved. ‘Low-resourced’-ness is a complex problem going beyond data availability and reflects systemic problems in society. In this paper, we focus on the task of Machine Translation (MT), that plays a crucial role for information accessibility and communication worldwide. Despite immense improvements in MT over the past decade, MT is centered around a few high-resourced languages. As MT researchers cannot solve the problem of low-resourcedness alone, we propose participatory research as a means to involve all necessary agents required in the MT development process. We demonstrate the feasibility and scalability of participatory research with a case study on MT for African languages. Its implementation leads to a collection of novel translation datasets, MT benchmarks for over 30 languages, with human evaluations for a third of them, and enables participants without formal training to make a unique scientific contribution. Benchmarks, models, data, code, and evaluation results are released at https://github.com/masakhane-io/masakhane-mt.

We describe a novel super-infrastructure for biomedical text mining which incorporates an end-to-end pipeline for the collection, annotation, storage, retrieval and analysis of biomedical and life sciences literature, combining NLP and corpus linguistics methods. The infrastructure permits extreme-scale research on the open access PubMed Central archive. It combines an updatable Gene Ontology Semantic Tagger (GOST) for entity identification and semantic markup in the literature, with a NLP pipeline scheduler (Buster) to collect and process the corpus, and a bespoke columnar corpus database (LexiDB) for indexing. The corpus database is distributed to permit fast indexing, and provides a simple web front-end with corpus linguistics methods for sub-corpus comparison and retrieval. GOST is also connected as a service in the Language Application (LAPPS) Grid, in which context it is interoperable with other NLP tools and data in the Grid and can be combined with them in more complex workflows. In a literature based discovery setting, we have created an annotated corpus of 9,776 papers with 5,481,543 words.

While the application of word embedding models to downstream Natural Language Processing (NLP) tasks has been shown to be successful, the benefits for low-resource languages is somewhat limited due to lack of adequate data for training the models. However, NLP research efforts for low-resource languages have focused on constantly seeking ways to harness pre-trained models to improve the performance of NLP systems built to process these languages without the need to re-invent the wheel. One such language is Welsh and therefore, in this paper, we present the results of our experiments on learning a simple multi-task neural network model for part-of-speech and semantic tagging for Welsh using a pre-trained embedding model from FastText. Our model’s performance was compared with those of the existing rule-based stand-alone taggers for part-of-speech and semantic taggers. Despite its simplicity and capacity to perform both tasks simultaneously, our tagger compared very well with the existing taggers.

Existing NLP models are mostly trained with data from well-resourced languages. Most minority languages face the challenge of lack of resources - data and technologies - for NLP research. Building these resources from scratch for each minority language will be very expensive, time-consuming and amount largely to unnecessarily re-inventing the wheel. In this paper, we applied transfer learning techniques to create Igbo word embeddings from a variety of existing English trained embeddings. Transfer learning methods were also used to build standard datasets for Igbo word similarity and analogy tasks for intrinsic evaluation of embeddings. These projected embeddings were also applied to diacritic restoration task. Our results indicate that the projected models not only outperform the trained ones on the semantic-based tasks of analogy, word-similarity, and odd-word identifying, but they also achieve enhanced performance on the diacritic restoration with learned diacritic embeddings.

Igbo is a low-resource language spoken by approximately 30 million people worldwide. It is the native language of the Igbo people of south-eastern Nigeria. In Igbo language, diacritics - orthographic and tonal - play a huge role in the distinguishing the meaning and pronunciation of words. Omitting diacritics in texts often leads to lexical ambiguity. Diacritic restoration is a pre-processing task that replaces missing diacritics on words from which they have been removed. In this work, we applied embedding models to the diacritic restoration task and compared their performances to those of n-gram models. Although word embedding models have been successfully applied to various NLP tasks, it has not been used, to our knowledge, for diacritic restoration. Two classes of word embeddings models were used: those projected from the English embedding space; and those trained with Igbo bible corpus (≈ 1m). Our best result, 82.49%, is an improvement on the baseline n-gram models.

Properly written texts in Igbo, a low-resource African language, are rich in both orthographic and tonal diacritics. Diacritics are essential in capturing the distinctions in pronunciation and meaning of words, as well as in lexical disambiguation. Unfortunately, most electronic texts in diacritic languages are written without diacritics. This makes diacritic restoration a necessary step in corpus building and language processing tasks for languages with diacritics. In our previous work, we built some n-gram models with simple smoothing techniques based on a closed-world assumption. However, as a classification task, diacritic restoration is well suited for and will be more generalisable with machine learning. This paper, therefore, presents a more standard approach to dealing with the task which involves the application of machine learning algorithms.