As part of the Open Language Data Initiative shared tasks, we have expanded the FLORES+ evaluation set to include Emakhuwa, a low-resource language widely spoken in Mozambique. We translated the dev and devtest sets from Portuguese into Emakhuwa, and we detail the translation process and quality assurance measures used. Our methodology involved various quality checks, including post-editing and adequacy assessments. The resulting datasets consist of multiple reference sentences for each source. We present baseline results from training a Neural Machine Translation system and fine-tuning existing multilingual translation models. Our findings suggest that spelling inconsistencies remain a challenge in Emakhuwa. Additionally, the baseline models underperformed on this evaluation set, underscoring the necessity for further research to enhance machine translation quality for Emakhuwa. The data is publicly available at this link.

Accepted to Main Conference

The accurate identification of loanwords within a given text holds significant potential as a valuable tool for addressing data augmentation and mitigating data sparsity issues. Such identification can improve the performance of various natural language processing tasks, particularly in the context of low-resource languages, that lack standardized spelling conventions. This research proposes a supervised method to identify loanwords in Emakhuwa, borrowed from Portuguese. Our methodology encompasses a two-fold approach. Firstly, we employ traditional machine learning algorithms incorporating handcrafted features, including language-specific and similarity-based features. We build upon prior studies to extract similarity features and propose utilizing two external resources: a Sequence-to-Sequence model and a dictionary. This innovative approach allows us to identify loanwords solely by analyzing the target word without prior knowledge about its donor counterpart. Furthermore, we fine-tune the pre-trained CANINE model for the downstream task of loanword detection, which culminates in the impressive achievement of F1-score of 93\%. To the best of our knowledge, this study is the first of its kind focusing on Emakhuwa, and the preliminary results are promising, as they pave the way to further advancements. We make our loanword dataset and source code publicly available to foster further research.

Stopword lists, an essential resource for natural language processing and information retrieval, are often unavailable for low-resource languages. Creating these lists is time-consuming and expensive, making automated stopword detection a viable alternative. This paper introduces a novel stopword detection approach that exploits the topological properties of co-occurrence networks to identify function words. By leveraging the connectivity patterns of function words in these networks, the proposed approach aims to achieve higher precision compared to traditional frequency-based methods. To assess the effectiveness of the network-based approach, we constructed co-occurrence networks for Tetun and Emakhuwa (low-resourced languages), as well as English and Portuguese. We then compared the performance of this approach with traditional frequency-based methods. The results indicate that the network-based approach consistently outperforms traditional methods, with in-degree emerging as the most reliable indicator of function words. This finding suggests promising prospects for automatically generating stopword lists in other low-resource languages, paving the way for developing natural language processing tools for these linguistic contexts.

Africa, which is home to over 2000 languages from more than six language families, has the highest linguistic diversity among all continents. This includes 75 languages with at least one million speakers each. Yet, there is little NLP research conducted on African languages. Crucial in enabling such research is the presence of labeled datasets by native speakers. In this paper, we introduce 14 sentiment labeled Twitter datasets in 14 African languages (Amharic, Algerian Arabic, Hausa, Igbo, Kinyarwanda, Moroccan Arabic, Mozambican Portuguese, Nigerian Pidgin, Oromo, Swahili, Tigrinya, Twi, Xitsonga, and Yoruba) from four language families (Afro-Asiatic, English Creole, Indo European, and Niger-Congo). We describe the data collection methodology, annotation process, and related challenges when curating each of the datasets. We also build different sentiment classification baseline models on the datasets and discuss their usefulness.

Major advancement in the performance of machine translation models has been made possible in part thanks to the availability of large-scale parallel corpora. But for most languages in the world, the existence of such corpora is rare. Emakhuwa, a language spoken in Mozambique, is like most African languages low-resource in NLP terms. It lacks both computational and linguistic resources and, to the best of our knowledge, few parallel corpora including Emakhuwa already exist. In this paper we describe the creation of the Emakhuwa-Portuguese parallel corpus, which is a collection of texts from the Jehovah's Witness website and a variety of other sources including the African Story Book website, the Universal Declaration of Human Rights and Mozambican legal documents. The dataset contains 47,415 sentence pairs, amounting to 699,976 word tokens of Emakhuwa and 877,595 word tokens in Portuguese. After normalization processes which remain to be completed, the corpus will be made freely available for research use.

This package provides a mechanism to automatically detect words in Emakhuwa borrowed from Portuguese.

The first spell checking tools for Emakhuwa Languanges and online Dictionary.

Emakhuwa is a Mozambican language under the low-resource category despite being widely spoken in Mozambique (i.e. over 6 million speakers). To the best of our knowledge, no Machine Translation tools exist for Emakhuwa . However, in recent years, there has been a huge col- laborative effort from African Natural Language research communities to develop techniques for Neural Machine Translation adequate to African Languages. This led to the development of corpora, text representation techniques, pre-trained models, and Neural Network Architectures, all of which are benchmarks for improving current Machine Translation of low-resourced languages, and in particular the African language family. Therefore, this study aims to investigate how this development can be helpful to assist Machine Translation of Emakhuwa, and also propose a suitable approach and resources to develop such system

Major advancement in the performance of machine translation models has been made possible in part thanks to the availability of large-scale parallel corpora. But for most languages in the world, the existence of such corpora is rare. Emakhuwa, a language spoken in Mozambique, is like most African languages low-resource in NLP terms. It lacks both computational and linguistic resources and, to the best of our knowledge, few parallel corpora including Emakhuwa already exist. In this paper we describe the creation of the Emakhuwa-Portuguese parallel corpus, which is a collection of texts from the Jehovah's Witness website and a variety of other sources including the African Story Book website, the Universal Declaration of Human Rights and Mozambican legal documents. The dataset contains 47,415 sentence pairs, amounting to 699,976 word tokens of Emakhuwa and 877,595 word tokens in Portuguese. After normalization processes which remain to be completed, the corpus will be made freely available for research use.