Abstract
It is important to be able to analyze the emotional state of people aroundthe globe. There are 7100+ active languages spoken around the world andbuilding emotion classification for each language is labor intensive.Particularly for low-resource and endangered languages, building emotionclassification can be quite challenging. We present a cross-lingual emotionclassifier, where we train an emotion classifier with resource-rich languages(i.e. \textit{English} in our work) and transfer the learning to low andmoderate resource languages. We compare and contrast two approaches of transferlearning from a high-resource language to a low or moderate-resource language.One approach projects the annotation from a high-resource language to low andmoderate-resource language in parallel corpora and the other one uses directtransfer from high-resource language to the other languages. We show theefficacy of our approaches on 6 languages: Farsi, Arabic, Spanish, Ilocano,Odia, and Azerbaijani. Our results indicate that our approaches outperformrandom baselines and transfer emotions across languages successfully. For alllanguages, the direct cross-lingual transfer of emotion yields better results.We also create annotated emotion-labeled resources for four languages: Farsi,Azerbaijani, Ilocano and Odia.