Applications of Neuro-Symbolic Artificial Intelligence

Prof. Boris Galitsky is a cofounder of Knowledge Trail, San Jose, CA, and teaches a course on LLM-based programming at Stavropol Agrarian University. He has contributed linguistic and machine learning technologies to Silicon Valley start-ups as well as to companies such as eBay and Oracle for over 25 years. His work in information extraction and sentiment analysis supported several acquisitions, including Xoopit by Yahoo, Uptake by Groupon, LogLogic by Tibco, and Zvents by eBay. His security-related document analysis technologies also contributed to the acquisition of Elastica by Symantec.As an architect of the Intelligent Bots project at Oracle, he developed a discourse analysis technique for dialogue management and documented this work in his book Developing Enterprise Chatbots. He also authored a two-volume monograph, AI for CRM, based on his experience developing Oracle Digital Assistant. In addition, he is an Apache committer to OpenNLP, where he created the OpenNLP Similarity component, a foundation for semantically enriched search engines and chatbot development.Prof. Galitsky’s research into the exploration and formalization of human reasoning culminated in his book Computational Autism, widely used by parents of children with autism and rehabilitation specialists. His focus on medical applications has resulted in two further research monographs: Artificial Intelligence for Healthcare Applications and Management, co-authored with Dr. Saveli Goldberg, and Health Applications of Neuro-Symbolic Artificial Intelligence.

• Adversarial integration of LLM and logic program.- Adversarial abductive dialogue framework with reinforcement for tackling LLM hallucination.- Improving ToM capabilities of LLMs in applied domains.- Rehumanizer: Enhancing LLM recommendations for human-like engagement and intriguing interactions.- A copilot for social promotion.- Abduction-based personalized recommender for Web 3.0.- Creating ad sequences with optimal discourse structure by LLM.- A computational framework for analyzing and supporting invention dialogues.- A system for distributed chemical engineering simulations.- Step wise approximation of CBOW reduces hallucinations in tail cases.