A Survey of Quantum-Cognitively Inspired Sentiment Analysis Models

Quantum theory, originally proposed as a physical theory to describe the motions of microscopic particles, has been applied various non-physics domains involving human cognition and decision-making that are inherently uncertain exhibit certain non-classical, quantum-like characteristics. Sentiment analysis is typical example such domains. In last few years, by leveraging modeling power quantum probability (a non-classical stemming from mechanics methodology) deep neural networks, range novel quantum-cognitively inspired models for sentiment have emerged performed well. This survey presents timely overview latest developments in this fascinating cross-disciplinary area. We first provide background at theoretical level, analyzing their advantages over classical theories cognitive aspects analysis. Then, recent introduced discussed detail, focusing on how they approach key challenges task. Finally, we discuss limitations current research highlight future directions.