This study aims to present an approach for the challenges of working with Sentiment Analysis (SA) applied to news articles in a multilingual corpus. It looks at the use and combination of multiple algorithms to explore news articles published in English and Portuguese. It presents a methodology that starts by evaluating and combining four SA algorithms (SenticNet, SentiStrength, Vader and BERT, being BERT trained in two datasets) to improve the quality of outputs. A thorough review of the algorithms’ limitations is conducted using SHAP, an explainable AI tool, resulting in a list of issues that researchers must consider before using SA to interpret texts. We propose a combination of the three best classifiers (Vader, Amazon BERT and Sent140 BERT) to identify contradictory results, improving the quality of the positive, neutral and negative labels assigned to the texts. Challenges with translation are addressed, indicating possible solutions for non-English corpora. As a case study, the method is applied to the study of the media coverage of London 2012 and Rio 2016 Olympic legacies. The combination of different classifiers has proved to be efficient, revealing the unbalance between the media coverage of London 2012, much more positive, and Rio 2016, more negative.

This investigation investigates the influence of a variety of data augmentation techniques on sentiment analysis in low-resource languages, with a particular emphasis on Bulgarian, Croatian, Slovak, and Slovene. The following primary research topic is addressed: is it possible to improve sentiment analysis efficacy in low-resource languages through data augmentation? Our sub-questions look at how different augmentation methods affect performance, how effective WordNet-based augmentation is compared to other methods, and whether lemma-based augmentation techniques can be used, especially for Croatian sentiment tasks. The sentiment-labelled evaluations in the selected languages are included in our data sources, which were curated with additional annotations to standardise labels and mitigate ambiguities. Our findings show that techniques like replacing words with synonyms, masked language model (MLM)-based generation, and permuting and combining sentences can only make training datasets slightly bigger. However, they provide limited improvements in model accuracy for low-resource language sentiment classification. WordNet-based techniques, in particular, exhibit a marginally superior performance compared to other methods; however, they fail to substantially improve classification scores. From a practical perspective, this study emphasises that conventional augmentation techniques may require refinement to address the complex linguistic features that are inherent to low-resource languages, particularly in mixed-sentiment and context-rich instances. Theoretically, our results indicate that future research should concentrate on the development of augmentation strategies that introduce novel syntactic structures rather than solely relying on lexical variations, as current models may not effectively leverage synonymic or lemmatised data. These insights emphasise the nuanced requirements for meaningful data augmentation in low-resource linguistic settings and contribute to the advancement of sentiment analysis approaches.

Natural language processing for languages with limited resources is hampered by a lack of data. Using English as a hub language for such languages, cross-lingual sentiment analysis has been developed. The sheer quantity of English language resources raises questions about its status as the primary resource. This research aims to examine the impact on sentiment analysis of adding data from same-family versus distant-family languages. We analyze the performance using low-resource and high-resource data from the same language family (Slavic), investigate the effect of using a distant-family language (English) and report the results for both settings. Quantitative experiments using multi-task learning demonstrate that adding a large quantity of data from related and distant-family languages is advantageous for cross-lingual sentiment transfer.

Named-entity recognition and classification (NERC) is an essential natural language processing (NLP) task involved in many applications like interactive question answering, summarising, relation extraction, and text mining. Available NERC corpora follow different annotation schemes that vary in terms of formats and levels of complexity according to research requirements: from 1-level hierarchy annotations (e.g., “Person”, “Location”, and “Organisation”) to multi-level schemes. Inspired by the work of the Universal Dependencies framework in terms of a standard representation of parsed trees, we developed the universal named-entity recognition (UNER) framework, which consists of a multi-level NERC hierarchy and a corresponding workflow that parses data from Wikipedia and DBpedia, translating it to UNER annotations. This chapter presents the UNER hierarchy and its workflow for data extraction and annotation. The proposed process was used to generate an English corpus, which was evaluated qualitatively and quantitatively. Furthermore, seven strategies for annotation improvement were presented and discussed, showing that the usage of information from the Open Event Knowledge Graph (OEKG) can improve our dataset.

This paper presents a cross-lingual sentiment analysis of news articles using zero-shot and few-shot learning. The study aims to classify the Croatian news articles with positive, negative, and neutral sentiments using the Slovene dataset. The system is based on a trilingual BERT-based model trained in three languages: English, Slovene, Croatian. The paper analyses different setups using datasets in two languages and proposes a simple multi-task model to perform sentiment classification. The evaluation is performed using the few-shot and zero-shot scenarios in single-task and multi-task experiments for Croatian and Slovene.

Gaurish Thakkar, Nives Mikelic Preradovic, Marko Tadić. Proceedings of the 9th Workshop on Slavic Natural Language Processing 2023 (SlavicNLP 2023). 2023.