Rui Mao

Rui Mao is a Research Fellow at Nanyang Technological University. He received his Ph.D. in Computing Science from the University of Aberdeen. His research interests include computational metaphor processing, and cognitive computing. He and his founded company have developed the first neural network search engine (https://wensousou.com) for searching ancient Chinese poems by using modern language, and a system (https://metapro.ruimao.tech) for linguistic and conceptual metaphor understanding. He has published several papers as the first author in top-tier conferences and journals, e,g., ACL, AAAI, IEEE ICDM, Information Fusion, and IEEE Transactions on Affective Computing. He served as Area Chair in COLING and EMNLP and Associate Editor in Expert Systems, Information Fusion and Neurocomputing. 

MetaPro and Its Applications for Cognitive Analysis

Concept mappings were commonly used as a medium for cognitive analysis through psychological tests such as the word association test, thematic apperception tests, and the Rorschach test. However, these tests are time-consuming and costly due to their reliance on one-on-one interviews. To address this issue, we developed a tool, termed MetaPro to automatically parse concept mappings from metaphorical expressions. Metaphors frequently appear in everyday language, reflecting concept mappings between target and source domains. Thus, we can use MetaPro to obtain concept mappings from daily textual data and analyse the cognitive patterns of a large population. We have employed this method to study cognition in various domains, such as cognitive pattern analysis for depression patients, financial analysts, and public perception of different types of weather disasters.

Neurosymbolic AI for Personalized Sentiment Analysis

Sentiment analysis is crucial in extracting valuable insights from vast amounts of textual data generated across various platforms, such as social media, customer reviews, news articles, etc. Over the years, researchers and business professionals have worked hard to refine sentiment analysis algorithms, but there is a limit to how accurate any algorithm can be without considering personalization. In this work, we propose a framework for personalized sentiment analysis that performs automatic user profiling by modeling users based on different levels of personalization, before performing sentiment analysis. In particular, such framework leverages seven levels of personalization (from bottom to top), namely: Entity, to distinguish between humans and other intelligent agents; Culture, to take into account how different cultures perceive the same concept as positive or negative; Religion, to consider how specific religious beliefs may affect an individual’s opinion about certain topics; Vocation, to better gauge people’s opinion based on their job and education level; Ideology, to take into account political beliefs as well as social, economic, or philosophical viewpoints; Personality, to better classify certain concepts as positive or negative based on personality traits; finally, Subjectivity, to take into account personal preferences and experiences.