In addition to creating an ETD, students now need to become their own cataloguers. The tools being developed by the NDLTD will assist in the submission of a basic description of the ETD but there are skills to be developed and issues to be aware of.
Students will have already discovered that one of the greatest barriers to finding information is the difficulty of coming up with the right terminology. Lists of standardized subject heading terms, structured thesauri, and fielded searching have been created to remedy this problem.
Accurate metadata improves precision and increases the recall of the content of ETDs by using the same standardized terms or elements. However, even if common metadata elements are used, there is no guarantee that the vocabularies, the content of the elements, will be compatible between communities of interest. Students and researchers working within specialized areas sometimes forget that language and terms often have particular and precise meanings. Outside this field of interest, global searches can return too much of the wrong information.
Generating accurate metadata requires some of the basic skills of resource description and good practice in avoiding three language problems that cause poor precision:
- Polysemy: words with multiple meanings. For example, if we are searching for an article that discusses types of springs and their uses, we might retrieve articles on freshwater springs or on the season of spring, as well as on leaf springs, flat springs, or coil springs.
- Synonymy: words representing the same concept, although they may do it with different shades of meaning. Take the words 'ball,' 'sphere,' and 'orb,' or 'scuba diving' versus 'skin diving.' If we look for scuba diving, but the term used is skin diving, we will miss materials we might otherwise find. Good metadata should draw these materials together, despite their use of different synonyms.
- Ambiguity: If we return to our example of springs, we can see what differentiates these meanings is their context. It is unlikely an article on coil springs will also discuss water quality. The other words used in the article, and the processes described, will be entirely different. A search engine must understand the meaning, not just be able to match the spelling of a word, if it is going to differentiate between different meanings of the same word. A possible solution to this difficulty lies in the recent development of the notion of Application Profiles. Application Profiles provide a model for the way in which metadata might be aggregated in ‘packages’ in order to combine different element sets relating to one resource. This is a way of making sense of the differing relationship that implementers and namespace managers have towards metadata schema, and the different ways they use and develop schema. Students should investigate these developments within their community of interest.
See: Metadata: Cataloging by Any Other Name ... by Jessica Milstead and Susan Feldman ONLINE, January 1999
Available [on-line] http://www.onlineinc.com/onlinemag/OL1999/milstead1.html
See: Application profiles: mixing and matching metadata schemas Rachel Heery and Manjula Patel
Available [on-line] http://www.ariadne.ac.uk/issue25/app-profiles/
Next Section: Fulltext