User-Emotion Detection Through Sentence-Based Classification Using Deep Learning: A Case-Study with Microblogs in Albanian

摘要: Human emotion analysis has always stimulated studies in different disciplines, such as Cognitive Sciences, Psychology, and thanks to the diffusion of social media, it is attracting interests computer scientists too. Particularly, growing popularity Microblogging platforms, generated large amounts information which turn represent an attractive source data be further subjected opinion mining sentiment analysis. In our research, we leverage performed on micro-blogging texts postings Albanian language, enables use technologies monitor follow feelings perception people with respect products, issues, events, etc. Our approach tackles problem classifying a text fragment into set pre-defined categories therefore aims at detecting emotional state writer conveyed through text. order achieve this goal, perform comparative between classifiers, using deep learning other classical machine classification algorithms. We also adopt domestic stemming tool for language preprocess datasets used second round experiments. Experimental evaluation shows that produces overall better results compared methods terms accuracy. present findings related length being processed impact classifiers’