An approach to solve social problems by probabilistic modeling of real-world phenomena

Abstract

Real world phenomena are described by various IT systems and text data, and the application of probabilistic modeling technology for constructing a probabilistic model from such real-world big data is increasing. In this presentation, we explain our efforts to solve real-world problems using such a probabilistic model-based approach, collect real-world data, model from data, develop support/service systems using probabilistic models, and continuously evolve the systems. Discuss the process. Based on examples, developed the systems and actual projects, we will clarify the methodology for developing AI application systems that can solve real-world problems.