Another source I found was the work of Simon et al. at Carnegie Mellon, which I introduced in my book. I saw an introduction in a sales email that read, “We will explain in detail, with easy-to-understand demonstrations, the construction techniques of the ‘Knowledge Graph,’ which is indispensable for realizing ‘Agentic AI.'” It continued, “To realize ‘Agentic AI,’ which makes autonomous decisions and acts accordingly, a ‘Knowledge Graph,’ which shows the correlation of information beyond mere data retrieval, is essential.” The Ministry of Economy, Trade and Industry also emphasizes the need for business analysis and, as a basis, data standardization. Wait a minute, let me check by searching “Knowledge Graph and Frame Theory” on Google. Not only does it show a strong relationship, but it also indicates that it can be traced back to Simon et al.’s Logic Theorist, General Problem Solver. That makes sense. CACM’s explanation of the Knowledge Graph includes Is-a link etc. It’s perfectly natural that there are high expectations for OpenText.
And that’s not all; the implementations and goals of ConceptNet and CYC are also synchronized to the cutting edges. Thus, with the advent of the Age of Artificial AI, the mainstream of AI, cultivated over more than 50 years, finally becomes the mainstream, supplementing the shortcomings of previous approaches with DNNs, and this will become the next generation. A groundbreaking convergence!
In my book, “A Narrative History of Artificial Intelligence” (Springer 2024), I have written extensively about the work of Simon et al., in Section 1.7, *Footsteps of Herbert Simon*, pp. 29-34. It also includes the explanation for Frame Theory and ConceptNet.