Re: [TaxoCoP] Validity Check
What you are describing is a faceted approach, which is very useful for browsing structures and search refinement. While Ranganathan discovered a method to describe a wide range of books, his categories do not map perfectly onto the modern world. You have to be particular about the characteristics that are useful to your usersThink Amazon: object is classified by medium (book, movie, music) with subtype mediums (dvd, bluray, cd, digital, etc), by genre,(books = fiction, non fiction, sci fi ) again with subgenres, by audience (generally age groups), by author/muscian/director. By combining components, you can reduce a very large group of data to managable parts.I have seen increasing numbers of E-commerce and large content websites using this or similar approaches.However, i think that the genetics analogy doesn't really fit. I agree with Nick that genetic permutations are more narrow and acutally they are more firmly heirarchical than many other systems. You would need to go a bit further to see a multi-hierarchical approach. For a authoritative look at genetic heiraries -- look at MeSH: http://www.nlm.nih.gov/mesh/MBrowser.html. They classify by organism and cross reference to diseases. Some of the diseases/disorders belong to multiple heirarchies:However, it is still more hierarchical than what you seem to be describing.Kelly Green
From: Nick Berry <infoglutton@...>
To: "TaxoCoP@yahoogroups.com" <TaxoCoP@yahoogroups.com>
Cc: "TaxoCoP@yahoogroups.com" <TaxoCoP@yahoogroups.com>
Sent: Sun, April 4, 2010 12:48:55 PM
Subject: Re: [TaxoCoP] Validity Check
Isn't that sort of what Ranganathan discovered? A way to break information down into it's most basic components. Except then you have names for each type; ie object=tree or person=John O'Gorman. The fact that each name could be almost limitless will defy the utility of such a descriptive system. With genes, types are limited - that is, there are x number of genes in the world, and they don't have individual expression.For the Tweet example, my friend just posted "Radiohead never gets old.". We could say object(type= musical band, individual expression=Radiohea d) + time(type=measureme nt; expression=never) + action(type= development; expression= becoming; tense=present) + characteristic( type=modifier; expression= age)....but this has far more potential permutations than the genetic code and attempting to classify all knowledge using componentry like this loses its functionality somewhere around O (n^3) ... just doesn't seem worth the effort.
On Apr 3, 2010, at 11:01 AM, "John O'Gorman" <jogorman@tiberon- ia.com> wrote:Good afternoon, all;As some of you know, I am writing a book - to be released in October - about my Q6 model for organizing information. There have been a number of contributing factors to this happening, and this group being one of them I was wondering if I could impose upon all of the posters and pundits out there to comment on an idea.There are two complementary concepts in genetics - genotype and phenotype - that I want to apply in a metaphorical way to the classification of digital assets. Basically a genotype is the sum total of an organism's genetic material, while a phenotype is the expression of the form, life cycle and behaviour of said genotype. I'm wondering if a case can be made for a relatively small set of base genotypes forming the foundation for all of the combinations and permutations we see in the digital world.The question is this: if a person could identify the primary base genotype - and the one or more genotypes that contribute to the end product (the phenotype) - is there a case to be made for a universal taxonomy complemented by "specialization- by-combination" ? In other words, does the group support the notion of a sort of digital DNA approach to classification?