Step 2: Facet Analysis

The next step when creating a thesaurus is to complete a facet analysis (Appendix 2). I found this to be the most difficult part of the process. For this step, I needed to place the original indexable terms which I pulled from the fifteen subject statements under very broad concepts. The result is supposed to be a very broad picture of how to group the terms together.

Faceted classification is important in the thesaurus construction process because it helps to analyse the subject fields and determine relationships between concepts. S. R. Ranganathan developed a system of faceted analysis that involved five categories called PMEST: personality, matter, energy, space, and time (Aitchison et al., 2000). This concept has been developed and evolved over time but faceted classification still contains those primary areas.

While doing my facet analysis, I found some of the categories to be fairly easy to understand the types of terms which would fit under that heading. For example, I found the term small-scale obviously fit under attributes. There really was no debate in my mind where that term should go. Also, for the most part, I found the terms under artefacts to be relatively easy to compile. The trouble with the facet analysis came when I had to determine the differences between the concepts of agents and patients, and process/functions and operations. Another problem I had when creating the facet analysis was leaving the headings very broad. I found it difficult to not break up the terms into more specific facets. For example, under Agents, I split the terms up between institutions and people/groups, however, I automatically wanted to split those groups up and under institutions have two categories, libraries and schools. I also originally had people/groups split up into library users and students. Breaking the groups down into further divisions was a step I was supposed to leave until the relationship construction but I found it hard not to do that right away.

There were also some terms that just did not seem to belong anywhere. I had trouble placing the terms policies, protocols, and standards. I originally put them under abstract entities because I did not know what else to do with those terms. Eventually, after re-reading my notes a few times I discovered there was a heading I had missed. Under entities by function I found the facet end-products. I found this to be a good location for these terms because they are all the end-products of some sort of evaluation process. I also ended up placing mission statements under this facet as well because it is also the end-product of evaluation.

I also had a lot of trouble with the two terms bibliometric analysis and discourse analysis. I placed them in several different places throughout the process of the facet analysis. For a long time I had them coupled with statistics. In my mind, these three terms should all go in the same place because I believed they all related to research in some way. Originally all three terms were placed in abstract entities but I was never really happy with that because although I thought statistics would fit there bibliometric analysis and discourse analysis just did not seem to be abstract. In the end, I placed bibliometric analysis and discourse analysis under operations because I felt as though they were both external activities. I moved statistics from abstract entities to end-products because it is the result of research. After a few rough drafts I was finally able to produce a facet analysis in which I was happy with the location of each of the terms.


Previous   |   Next