Amazing technological breakthrough possible @S-Logix

Office Address

  • #5, First Floor, 4th Street Dr. Subbarayan Nagar Kodambakkam, Chennai-600 024 Landmark : Samiyar Madam
  • +91- 81240 01111

Social List

Research Proposal in Utterances and Emoticons based Multi-Class Emotion Recognition

Research Proposal in Utterances and Emoticons based Multi-Class Emotion Recognition

  Emotion recognition is the budding research area that utilizes technology to identify human emotions. The goal of emotion recognition is to classify a user-s current emotional state using input data automatically. The need for emotion recognition is to improve the human-machine interpersonal relationship. Different classes of emotions are reflected from user input data, such as anger, disgust, fear, guilt, joy, sadness, shame, and many more.

  Multi-class emotion recognition is developed to classify different classes of emotions. Multi-class emotion recognition categorizes each input data from multiple target labels based on the trained knowledge of multiple classes. The most powerful factors to identify the emotional state of the user are utterances and emoticons. Utterances denote the small comment or spoken word that expresses something aloud, and emoticons represent facial expressions to convey the user-s feeling. Emotion recognition based on utterances and emoticons enables a better understanding of user sentiments from their conversation. Multi-class emotion recognition based on utterances and emoticons effectively identifies user-s multiple states of emotions with high accuracy.

  • Emotion analysis is emerged as a trending research topic in natural language processing, mainly concentrating on a broad scope of human emotions and sensitivities.

  • Multi-class emotion analysis by utilizing contexts of utterances and emoticons is applied for sentiment analysis-related applications.

  • A recent approach to predict emotion through emotion analysis of conversational texts is evolved.

  • Prediction of emotion from multi-turn textual utterances is implemented using embedding and deep learning models in order to achieve high precision.

  • More recently, depression detection has been incorporated with multi-label emotion analysis to investigate the learning emotion features from several depressive-related factors.

  • External knowledge information is infused to perform better for the multi-label emotion classification task on depressive suicide notes.