Public knowledge intelligence
Mzizi
Public information is only useful when you can tell what is true, when it was true and where it came from.
Mzizi explores civic and political knowledge as a structured system: people, institutions, claims, records, dates and sources connected in ways that can be inspected instead of hidden behind a generated answer.
Knowledge model
Answers need a history.
Political and civic information changes over time. A useful system needs to model both the fact and its validity window, then expose the evidence behind it.
Time
Facts are attached to periods and events instead of treated as permanently true.
Provenance
Every important claim can point back to records, documents or identifiable sources.
Relationships
People, institutions, positions and events form an inspectable graph of public knowledge.
Retrieval
Search and AI interfaces should show why a result was returned, not only produce prose.
Example provenance chain
From claim to evidence.
Research questions
Trust has to be engineered.
How should conflicting public records be represented instead of silently reconciled?
How can a system distinguish current facts from historically correct but outdated facts?
What evidence should be visible to a user before an AI-generated civic answer is considered trustworthy?
How do we make political knowledge searchable without flattening uncertainty or context?