Skip to content

README.md

Latest commit

 

History

History
22 lines (17 loc) · 926 Bytes

README.md

File metadata and controls

22 lines (17 loc) · 926 Bytes

metaparse

This is to be a machine learning project to extract metainformation about information sources, such as agents online that share information.

Problem specification

Given some data from an information source I, recognize its intent about content towards the world, answering following questions about the information source and its information content:

  • F(domain).assets:
    • .Agent (who?)
    • .Place (where?)
    • .Event (when?)
    • .Topic (about what?)
  • X(process).actions:
    • . doing (doing what?)
  • Y(range).targets:
    • .Goal (why?)
    • .Idea (how in principle?)
    • .Plan (what specifically?)
    • .Step (in what order?)

This will allow to treat every information source as an instance of equation optimization process F(X)=Y, where the actions of agents are explained by their intents (Y) about content their world model (F).