E1: There is a gap in all elementary thoughts and/or actions. It is not known what will happen as a whole, based on an elementary action or construction of an element. Hence the importance of pattern recognition by humans.
E2: The events that make up the modeled knowledge base are generally not considered! But without the effective knowledge that gives it meaning, it is not possible to balance the load to execute the algorithms correctly!
E3: This is, in our understanding, the main problem with the applied use of ontologies! Terms far from people's reality! As in the case of health, a large hospital that cannot adhere to an ontology like Snomed! Main comment from doctors and nurses: "This ontology is very complete! Too much! It is not part of the reality of our daily diagnostic process!"
E4: When we apply pattern recognition techniques, we have classification analyses with abstract terms! Ok, the patient has cancer, but the doctor cannot say that it is because the neural network suggested it! Critical point: "Explainable AI versus inexplicable AI!" The challenge in this field is to make inexplicable AI as explainable as possible! Excuse the pun, it was irresistible!
E5: After ten years teaching object-oriented programming! Seeing a methodology that removes the method (process) from the entity (class) changes one's perspective! In this methodology, Entities are squares and processes are circles! Circles are actions that define the level of abstraction and entities can be used by different processes. This methodology already has ISO and the MIT guys (Dov Dori) use it in aerospace development! Their idea is to correlate everything with everything, creating a co-occurrence matrix! Componentizing the entire system by creating small interconnected apps. Doing service-oriented development and modeling, leaving relational integrity within the components and integrating components through services.
E6: In the book Clean Code we have a very well-founded criticism that non-componentized services generate more problems than solutions in a software development project! E7: It turns out that this fact (AI vs Models) proposes an integration between the conceptual universe and the classification of events (instances). From there, using Von Neumann's idea of integrating automata and conventional machines! Unfortunately, he did not complete this challenge before he died prematurely! We could use the ideas of the complex systems people! (Douglas Hofstadter and Melanie Michell). Automata are easily modeled! But it is impossible to predict their behavior on a large scale... As an example, this is why pattern recognition, such as liver cells, may not be working!!! It is much easier to identify the symptoms of yellow eye in the patient!
E8: From there, we can see that modeling is on one end and pattern recognition on the other! But if we put Sync Schema in the middle as a reference model for knowledge synthesis! The technology that comes closest to this is Graph DB, which unifies the universes for this possibility! The model would have to validate the universe by understanding the classification need! On the other hand, the classification would have to be explained... Proposing the creation and definition of strategic concepts... initially defined in an abstract way by AI... should be validated by experts to balance the base...
E9: It is observed that today we make queries based on graphs, regardless of the content! In them we can identify the PageRank, just by the star shape of the nodes with many edges (relationships)... something like: select all nodes with n relationships of this type and m of this other type... we can recognize complex design patterns in these structures (example snowflake, star and everything else...) .. we write the queries in Cypher... which is a kind of SQL for Graph DB developed by Neo4J. We can create properties, relationships, and synthetic nodes within a GDS (Graph Data Science) script. We often use the example of an Indian who created virtual airports between America and Europe. His challenge was to find the best airfare to two distant countries. In this example, relational databases broke and Graph DB delivered the shortest path in a simple and fast way. Studying the cognitive behavior of doctors in the diagnostic process for a long time now. It can be said that human beings work at the limit of mathematics without numbers. A doctor can say what a patient has. But he does not know how to correlate the diagnostic events of all the patients he has treated in his life! He simply does this unconsciously!
E10: “You cannot know what will happen in a battle, because there are so many events that it is impossible to predict them.” One of the author's main messages is incorporated into his masterpiece, War and Peace, Leo Tolstoy.
