Challenges, issues and future direction for text-based emotion detection in Natural Language Processing (NLP)

Torres-Cruz F.; Herrera-Urtiaga A.P.; Coyla-Idme L.; Cauna-Huanca G.J.; Lourens M.E.; Verma D. · AIP Conference Proceedings (2024) · DOI: 10.1063/5.0218218

Resumen

In today's society, communication takes place over a large geographic region, thanks to the expansion of the Internet. Textual communication and human contact take place on social media sites, apps, and other platforms. Emotion is an important part of a person's day-to-day existence since it allows them to express their feelings and situation. Speech, gestures, facial expressions, and writing may all be used to communicate emotions. From the study of emotions, numerous domains including data mining, psychology, human-computer interaction, neuroscience, e-learning, cognitive science and information filtering systems can get benefited. Many elements related to text mining including emotions may be investigated using the large volume of text available on consumer reviews, blogs, social media and news items. Emotion recognition in addition to analysis is one of the most interesting and evolving challenges in Natural Language Processing (NLP). This work offers an outline of rapidly expanding subject of emotion detection, with an emphasis on the current generation of social media platform-based algorithms as well as the obstacles. The objective of this work is to provide a narrative of the available techniques, models and restrictions for detecting emotions from text that researchers may use when performing emotion detection studies. Finally, the study discusses various unresolved challenges as well as future research directions in text-based emotion recognition.

Instituciones: Universidad Nacional del Altiplano