Subject: RE: analysis methodology
Newsgroups: gmane.comp.infodesign.facetedclassification
Date: 2002-12-19 12:37:46 GMT (6 years, 28 weeks, 2 days and 57 minutes ago)
> > Excellent! I actually posed a question along that line when I first drafted > my previous email. And in my blissful ignorance, assumed that they were > regarded as fine examples but had reservations as I stated with epicurioius. > So, what are these sites lacking? Incomplete or improper facet analysis > resulting in a less than desirable classification? epicurious, yes, you and Seems to be an attempt at facet analysis that was undertaken with an incomplete understanding of several principles. > I agree. How about wine.com? Well, there are categories here.... But are they facets? Discussion of the site tends to indicate that the facets are only one level deep. What is the facet? Personality? Flavor? Popular notions of wine characteristics. In any case (if memory serves correctly, you can't combine each feature, for searching either). > > What methodology did you use to define these isolates? To me, they make > sense...they feel right. Unfortunately, that doesn't agree with the folks > who ultimately sign off on it. Ah, the thousand dollar question. I didn't do several things and answered your question on the fly, it was an attempt at dividing the ingredient facet into logical areas. NOTE: there is not one right way to do this. That is why context is so vitally important. (And another quite long answer to your questions follows). The methodology should include an approach in which you are able to identify your universe of entities, it is best when beginning facet analysis to work within a bounded universe. I give students in my class a bag of marbles. Who is your user group? Why do they want this ordering/access system? What will they be using it for? Collectors who want to purchase marbles? Kids who want to play with them? Artists who want to illustrate a book about them? These considerations will affect your facets. At Epircurious we are dealing with recipes-so the best approach to conducting facet analysis is to get a good sense of the domain area. (This is one stepwhich I did not do, but rather relied on my own knowledge of cooking and recipes - something I would NOT recommend that you do if conducting facet analysis.) Here I'd recommend working with several comprehensive cookbooks in order to get a sense of how the domain is structured. For web design with a site like epicurious, you'd also be laboring (I assume) with knowledge of who is supporting the site (or who would be potential advertisers) and who your potential audience might be. Next step would be to take a look at a good glossary of cooking terms, and to familiarize yourself with a number of recipes themselves. Each step, examination of the cookbooks, the terms used in a glossary and the recipes themselves brings to light the detailed structures of the domain. (Channelling Vickery again, this time p. 20 on). Facet analysis is conceptual analysis of the subject matter, but being aware of the literature or domain will keep you on track, and make certain that all important and significant characteristics of the domain are covered. Even so constrained a universe as one composed entirely of recipes, has a virtually unlimited number of characteristics by which the domain could be divided. > > In Ranganathan's scheme of classification, he suggests that there are five > fundamental categories: personality (distinguishing characteristic), matter > (physical material), energy (action), space (location), time (time period). > > Now one question I've always had is whether one should look at these > categories and find the facets that naturally make sense within them. Or is > it a mistake? Should one do the analysis without Ranganathan's scheme and in > the end discover that yes, the facets we end up defining organically fit his > fundamentals? > > To me, doing it without Ranganathan's scheme makes sense...giving the > analysis the most unbiased view possible. However, would I be missing out on > a valuable tool? The members of the Classification Research Group in London in the 1950's worked extensively with Ranganathan's formulations of the five fundamental facets. They found that a longer list of fundamental categories proved helpful for their purposes (creating a number of faceted schemes for special subject areas). Substance, organ, constituent, structure, shape, property, object of action (patient, raw material), action operation, process, agent, space and time. You may find while working in a specific domain that you have other fundamental categories, or do not have a number of these. That is fine. The idea here is that the categories are recognized and distinguished. Not everyone finds having a number of fundamental categories a useful device for facet analysis. Ranganathan's fundamental categories are a bit mysterious for some. Vickery again, "Of these, Space and Time are relatively straightforward geographical and chronological schedules. Energy covers categories such as problem, method, process, operation, handling and technique. Matter comprises constituent materials of all kinds. Personalities (the most difficult for many to grasp) include libraries, numbers, equations, wavelengths of radiation, engineering workds, organisms, crops, religions, art styles, literary workds, languages, social groups, communities and states... Despite the generality and vagueness of these terms, more precise categories can often be assimilated to them to a surprising extent. Other lists of fundamental categories from the 1960's: Jesse Shera and Margaret Egan (US) agent, act, tools, object of action, time, space and product. Barbara Kyle (UK) natural phenomena, artefacts, activites, and "purposes, aims, ideas and abstracts" DeGrolier (FR) "constant categories" time, space, action, "variables" substance, organ, analytic, synthetic, property, form and organization. I add BOLD EMPHASIS to the following statement by Vickery (p. 24) Any such list of fundamental categories should not be used mechanically and imposed upon the subject, but to use it as a provisional guide in approaching a new field can be helpful. It provides an outline framework which may fit the field and give guidance in suggesting possible characteristics which should not be overlooked. It must never be taken to exhaust the field, or to be necessarily applicable in all subjects. _________ Think of fundamental categories and facet analysis as a swiss army knife sort of tools. Use them carefully and wisely and use the proper blade for each application. Facet analyusis is an inductive approach to knowledge organization: In sum these are the premilimary steps: Identify all entities (things concepts etc.) by analyzing the domain This results in a big pot of disorganized entities. These are your isolates. This is the analytical step in what is called an analytico-synthetic approach. Then you must group the isolates together by dividing them according to a single characteristic. To use a very common example: Education we group the concepts Primary, Secondary, Further together because these isolates are all distinguished by the Age characteristic. Geography, Mathematics and Physics are all distinguished by the Subject taught characteristic. approach). I was a bit loose with my terminology before, and there are better more approachable terms to use, but these are the most common in the literature. All isolates that share a particular characteristic constitute a facet. All isolates, once grouped together become foci (since they are now organized.) Again a long answer to your question. Please let me know if this was unclear, I just threw a lot of material your way. Kathryn http:/ella.slis.indiana.edu/~klabarre/portfolio.html To unsubscribe from this group, send an email to: facetedclassification-unsubscribe <at> yahoogroups.com Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/