Research corpus

Notes on Kenyan citation paths

The research index collects Kijito Citation Lab's reviews of AI answers about Kenyan businesses, categories, counties and evidence types. Materials include source-path breakdowns, English–Swahili comparisons, registry-evidence notes, platform-proxy cases and entity-collision reviews. New materials appear when a topic has enough repeated observations to be worth publishing, rather than on a fixed editorial calendar, so each note can keep its claim, citation and support level attached.

Four source roles inside an AI answer · en

Direction I

Source paths and platform proxies

Which sources are allowed to speak for a Kenyan business: Kenyan-owned pages, official local records and county references set against maps, social profiles, directories and international aggregators. Local record, local story, platform proxy and unsupported echo.

Direction II

English–Swahili retrieval paths

How the cited source changes when the same Kenyan business, category or location query is asked in English and in Swahili. Language-sensitive cases where a phrasing shift moves source choice, entity match or claim strength.

Direction III

Registry evidence and entity collisions

Whether structured local records, tax identifiers, licence references and trade-body mentions are treated as citation support, and how business-name collisions across counties, sectors and platform profiles push answers toward the wrong match.