When I was researching the …grain of salt post I happened across the entry for sodium chloride at Wikipedia.
I don’t know how many times I have looked at Wikipedia pages but that day I noticed the headings in the sidebar that read:
IUPAC name (International Union of Pure and Applied Chemistry nomenclature)
Other names
Identifiers
Properties
Structure
Hazards
Related Compounds
Supplementary data page
Think about it for a minute.
Substances don’t arrive in labs, say for example the fictional labs seen on CSI with IUPAC names, other names, or even identifiers.
How are they identified? Can you say by their properties?
Now there is an odd dis-connect between indexing and identification.
That is indexing is by names and identifiers, both of which are known to be weak, rather than by properties.
Now there is an idea, an indexer that marshals properties for any index entry and can report why a particular entry was made.
We would not accept any less from a lab analysis, I wonder why we accept it from our indexers?
Subjects, other than substances, also have properties, including relationships to other subjects.
Identifiers and locators in topic maps are quick and convenient ways to navigate topic maps and the subjects represented therein.
We should now allow that convenience to blind us to the deeper complexity of reliable identification of subjects by their properties.
Indexing based upon more than names and identifiers looks like a largely unexplored landscape and one where topic maps could make an original contribution to the art of indexing.
Well, to be honest, topic maps would be making explicit what indexers have been doing for years. Which would make it even more valuable.
Indexing by Properties. Has a nice ring to it doesn’t it?
Has a number of implications for semantic web technologies, but more on that anon.