Interpretable AI for applications

Course provides practical insights for building applications involving elements of social intelligence technology involving graph analysis, semantic modeling, natural language processing and approaches for developing natural language comprehension and production systems, bot automation, social networks and transactional networks such as blockchains and marketplaces.
First block of the course is dedicated to different aspects of social data analysis and graph analysis in particular. It provides an explanation of different sorts of graph structures, applications of the graphs for semantic modeling, support of the graph processing in different systems and projects, with discussion of alternative approaches for graph representations. Also, the issue of explainability of the machine learning technology is discussed, including theoretical aspects of symbolic and sub-symbolic artificial intelligence and practical requirements to the modern machine learning solutions (GDPR, XAI, human-friendly AI).
Second block provides theoretical and practical aspects of building applications for social computing such as analytics in social networks or building applications for these networks including traditional centralized social networks such as Facebook and VKontakte as well as decentralized social and financial networks such as Ethereum, Steemit and Golos. Also, the cybernetic aspects of democracy are explained along with practical implications for building recommendations and reputation systems.
Third block of the course covers the practical aspects of developing natural language applications used to capture semantics and sentiment in online communications as well as production of texts in conversational systems such as chat-bots. Also, the notion of artificial general intelligence (AGI) is elaborated along with discussion about its feasibility and practical applications.