Social media contains a lot of emotional information. How to accurately and efficiently recognise the emotional polarity of different language texts is a major challenge in the field of natural language processing. Systematically identifying the emotional polarity of global users on specific topics or events is of great significance. However, different languages have different ways and degrees of emotional expression. Building a systematic language model to process multilingual text is a research challenge. This research focuses on the problem of knowledge gaps among languages and obtaining training data for each language. We propose a multilingual emotion analysis framework based on shared, multilingual, emotion word embedding and a pre-trained language model that realises the emotion polarity prediction of multilingual texts in the context of one language training data. An accuracy of 0.92 was achieved through tests on 8 different target language texts, which proved the feasibility and efficiency of this framework.
Li, Yuming; Chan, Johnny; Peko, Gabrielle; and Sundaram, David, "MULTILINGUAL SENTIMENT ANALYSIS IN SOCIAL MEDIA: CATCH THE EMOTION BEHIND DIFFERENT EXPRESSIONS" (2021). PACIS 2021 Proceedings. 69.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.