In my presentation, I will introduce state-of-the-art methods for textual and multimodal sentiment analysis and its specific sub-tasks, such as sentiment classification, aspect extraction, aspect-based sentiment analysis, and sarcasm detection, briefly addressing deep-learning and symbolic approaches. Further, I will explain the role of sentiment and emotion recognition in building an affective dialogue system that interprets and generates dialogue with emotional coherence. This requires understanding the sentiment and affect in conversations, which largely depends on such factors as context and the state of each speaker. In particular, I will introduce three important methods to detect sentiment and emotion in conversation based on the context, memories, and recursive neural networks. Finally, I will outline the shortterm and medium-term prospect for the development of the sentiment analysis field and suggest future work directions.
Soujanya Poria holds a Ph.D. degree in Computer Science from the University of Stirling, UK. He is a recipient of the prestigious early career research award called ‘NTU Presidential Postdoctoral Fellowship’ in 2018 which offers him a research grant worth US$150,000. Before taking up his presidential fellowship position at NTU, Soujanya was a scientist at the A*STAR and the Temasek Laboratory, NTU. He is also (co-)PI of multiple academic and industrial grants with the amount totaling to US$ 250,000. Soujanya has co-authored more than 60 papers, published in top-tier conferences and journals such as ACL, EMNLP, AAAI, NAACL, Neurocomputing, Computational Intelligence Magazine, etc.. He is also an adjunct faculty at Indraprastha Institute of Information Technology, Delhi, India and an adjunct scientist at A*STAR, Singapore. Soujanya served as a senior PC member at AAAI 2019, IJCAI 2019 and often serve as a PC member in reputed conferences such as ACL, EMNLP, IJCAI, NAACL. Soujanya has Google Scholar citations of more than 3500 and his h-index is 33. Recently, in an article published in the Journal of Information Sciences, Soujanya has been listed as one of the most prolific and impactful researchers from 2000 to 2016 in the field of sentiment analysis.