Emotion Detection of Contextual Text using Deep learning
Umar Rashid, Muhammad Waseem Iqbal, Muhammad Akmal Skiandar, Muhammad Qasim Raiz, Muhammad Raza Naqvi, Syed Khuram Shahzad
Abstract
In recent years, the data in the form of text is created in very huge amount in day to day conversation on social media. We need a naive approach to analyzed and summarized data to extract meaningful information. Textual dialogue is given in contextual emotion detection. We have to recognize user emotions either it is happy, sad, angry or others. This paper describes Aimens system which detect emotions from textual dialogues. This system used the Long short term memory (LSTM) model based on deep learning to detect the emotions like happy, sad and angry in contextual conversation. The main input to the system is a combination of word2vec and doc2vec embeddings. The output results are shown substantial changes in f-scores over the model baseline where Aimens system score is 0.7185.