Metadata based indexation , for contextual information
retrieval this approach relies on the context of creation of recorde, and
ideally on their physical arrangement reflecting it. The main finding tool is
the archival inventory. This is the mainstream approach currently followed to
build information systems for historical archives, and its affordance is
browsing, i.e., a set of archival units is selected via contextual metadata and
then perused thoroughly in search for specific information.
Entity-based indexation, for referential information
retrieval this approach relies on selected searchable contents (entities of
interest), for the purpose of finding possibly relevant documents. The main
finding tool is the index, linking entities with relevant documents (where they
are mentioned). Its affordance is a mix of searching (for entities) and browsing
(documents), typically on a much smaller search space than using the contextual
metadata approach.
Structured entity-based indexation, for information
extraction this approaches still focuses on selected contents (entities of
interest), but entails the extraction and creation of structured information
from the relevant documentation. Its affordance is searching, and the main
finding tools are databases.
Full content-based indexation, for information retrieval
these approaches rely on the full availability of contents, thus are by far and
large only possible in a digital setting. The affordance is searching, and the
finding tool is the (full-text) search engine. Google Books is an example of an
IR approach to search, on a collection of digitized objects.