We have completed numerous changes in the code of the Core Expert. Its structure has become simpler, and the possibilities are greater. In early December, we will present new features in the demo version.
We also created the basis for two new layers of memory (the 5th and 7th) and began to fill them with knowledge. These layers do not enter directly into the core and may be located separately from it. In addition, this knowledge can be shared among several cores.
Several modules have been developed that simplify textual and, in the future, speech interaction with the core. In the near future we will add the ability to correctly recognize words entered with typos or errors.
We have also almost completed new methods of extracting knowledge from the body of texts and adding them to the core. It remains to understand the morphology of unknown words to the core, representing geographical names and names.