ABOUT THE CONCEPT OF SEMANTIC MODELING Part 1. METHODOLOGY AND THEORY OF THE CONCEPT
Vitaly Shamilovich Gumirov1, Andrey Vitalyevich Gumirov2, Dmitry Ivanovich Sviridenko3
1Eyeline CIS, Novosibirsk, Russia
2Novosibirsk State University, Novosibirsk, Russia
3Institute of Philosophy and Law of the Siberian Branch of the Russian Academy of Sciences, Center for Artificial Intelligence, Novosibirsk National Research State University, Novosibirsk, Russia
Keywords: Computer science, mathematical logic, predicate calculus language, constructive model theory, semantics, methodology, programming theory, no-code and low-code, executable specifications, problem, problem solution criterion, problem solution context, semantic modeling, ontology, artificial intelligence
Abstract
This article opens a series of papers devoted to a holistic, comprehensive presentation of the concept of semantic modeling, including its methodological, logical-mathematical, and technological aspects, with an emphasis on practical applications and the concept's relationship with such areas of the IT industry as programming practice and artificial intelligence. For this reason, the central focus of this series of articles will be the description and discussion of the organizational and technological framework for the practical application of the concept's principles. A few clarifications are in order. Research on semantic modeling began in the late 1970s and early 1980s and is actively pursued today. The main focus of this research is the creation and development of a methodologically and theoretically sound technology for automatically solving intelligent problems as a trusted artificial intelligence technology, and the application of this technology and its tools to solving practical problems. The primary methodological and theoretical efforts of this research are aimed at creating, studying, and developing a declarative logical-mathematical formalism for problem description whose procedural semantics would allow for the automatic extraction of a problem-solving algorithm from its declarative specification and subsequent execution of this algorithm by a computer. The description of the methodological and theoretical results obtained here is the subject of this article. Naturally, the results obtained in this area should find practical application, which will be reflected in the second article in the series, dedicated to describing the semantic modeling technology created and being developed by the authors of this series. Since the problem of further improving this technology by, among other things, applying certain tools of modern artificial intelligence (AI), in particular, large-scale language models (LLM) and generative AI, is of undoubted interest, the third article in the series will be devoted to this issue. It should be noted that the first two articles in this series will focus on explaining the potential of semantic modeling as it applies to practical programming in the IT industry. The topic of deeply automating manual programming, particularly the coding stage, has recently been actively discussed in computer science. This topic has become especially popular with the advent of large language models. It turns out that semantic programming can also be successfully used to solve this problem. For this reason, the first two articles in this series will be devoted to discussing this potential. The relationship between semantic modeling and artificial intelligence will be the subject of the third article.
|
