Automatic Detection of Sentiment and Theme of English and Ukrainian Song Lyrics

Kryva, Uliana; Dilai, Marianna

doi:10.1109/stc-csit.2019.8929732

Cited by 5 publications

(3 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Several researchers have addressed this issue by exploring innovative approaches that leverage transfer learning, cross-lingual embeddings, and low-resource techniques to adapt existing NLP models to under-resourced languages. These efforts have shown promising results in various NLP tasks [3], [5], [7], [19], [20], such as part-of-speech tagging, named entity recognition, and machine translation.…”

Section: Overview Of Nlp In Under-resourced Languagesmentioning

confidence: 99%

“…Sentiment analysis in resource-limited settings has gained attention as researchers strive to overcome data scarcity and lack of linguistic resources. Approaches such as domain adaptation, unsupervised learning, and active learning have been explored to mitigate the limitations [14], [20], [21] imposed by small-scale datasets. Furthermore, resource-light techniques, including lexicon-based methods and feature engineering, have shown effectiveness in sentiment analysis tasks when extensive labeled data is not available.…”

Section: Sentiment Analysis In Resource-limited Settingsmentioning

confidence: 99%

See 1 more Smart Citation

Exploring Somali Sentiment Analysis: A Resource-Light Approach for Small-scale Text Classification

Bahar,

Ramaha

2023

ICAENS

View full text Add to dashboard Cite

Sentiment analysis, a fundamental task in natural language processing (NLP), plays a crucial role in understanding people's opinions and emotions expressed in textual data. While sentiment analysis has been extensively studied for major languages, under-resourced languages like Somali have received limited attention in this domain. This paper aims to address this research gap by proposing a resource light approach for sentiment analysis in Somali, which is tailored to the language's unique characteristics and limited linguistic resources. We present a methodology that combines lexicon-based methods and feature engineering techniques to effectively extract sentiment information from Somali text. A sentiment-annotated dataset was created through crowdsourcing, enabling the training and evaluation of a sentiment classification model specifically designed for Somali. Experimental results demonstrate the competitive performance of our approach compared to existing sentiment analysis techniques for under resourced languages. The findings highlight the feasibility of sentiment analysis in Somali, even with a small-scale dataset, and shed light on the implications for sentiment analysis in other under-resourced languages. This research contributes to the advancement of sentiment analysis capabilities for under resourced languages, empowering researchers and practitioners to gain insights from sentiment information in diverse linguistic contexts.

show abstract

Section: Overview Of Nlp In Under-resourced Languagesmentioning

confidence: 99%

Section: Sentiment Analysis In Resource-limited Settingsmentioning

confidence: 99%

Exploring Somali Sentiment Analysis: A Resource-Light Approach for Small-scale Text Classification

Bahar,

Ramaha

2023

ICAENS

View full text Add to dashboard Cite

show abstract

“…Celles-ci ont été utilisées dans de nombreux projets hors de France (par exemple : Meinecke, Hakimi, and Jänicke 2021 ;Baltazar and Västfjäll 2020 ;Kryva and Dilai 2019). Concernant le rap français, quelques chercheurs se constituent manifestement des corpus massifs Klimentová (2022), tout comme les blogueurs spécialisés (voir par exemple l'ambitieuse infographie de RapMinerz ( 2023)). Mais ces corpus dorment dans les disques durs de leurs collectionneurs, et, on l'espère, sont passés sous le manteau à l'occasion.…”

Section: Introductionunclassified

LRFAF : une exploration numérique du rap français depuis les années 1990

de Courson

2024

Preprint

View full text Add to dashboard Cite

Cet article introduit le grand corpus ouvert LRFAF, qui contient 37307 textes de rap français et leurs métadonnées. En utilisant des méthodes de lexicométrie et de traitement automatique des langues, il se propose une première exploration numérique du corpus. Il combine (i) une analyse diachronique, à la recherche des tendances lourdes connues par le rap français pendant ces trente dernières années, et (ii) une étude transversale, qui compare les rappeurs, à la recherche des lignes de force qui décrivent leur style. L'analyse diachronique révèle deux phases. Les textes deviennent plus noirs et plus agressifs entre 1990 et 2014, dans une sorte d'élargissement de la fenêtre d'Overton, avant de devenir plus légers, polis et moins politiques depuis, dans un mouvement de \enquote{variétisation}. L'analyse transversale aboutit à une analyse factorielle interprétable, qui propose de résumer le style des rappeurs par deux axes, que nous avons nommé littérarité et hardcorité.

show abstract