Sentiment analysis and opinion mining: Keynote address

Sentiment analysis and opinion mining are closely related tasks that recently have received great attention from both research community and industry. Opinion mining is aimed to turn a huge amount of user-contributed context in Internet, such as tweets, product reviews, and blogs into useful information, which allows the companies and the governmental bodies to improve their product and services, and, on the other hand, allows the consumers to make informed buying decisions. Sentiment analysis, apart from its use in opinion mining, finds other important applications in security, healthcare, and education, among others. In this paper, I briefly discuss the motivation behind these tasks and outline some techniques recently developed by our group, mainly based on specially tailored deep-learning architectures or other machine-learning methods.