Evaluation of Retrieval, UKDA
The UKDA database works in a very different manner from the BHI. The first difference that matters for this analysis is how the search results are displayed. In the BHI there is a choice in how the results appear. In the UKDA there is only on immediate results list. This list can work as both an alphabetical display and as a rotated display. The words are split apart and each has its own list. The term 'art' found the terms 'academic departments', 'Alliance Party of Northern Ireland', 'apartheid', 'apartments' and 'art'. Searching 'history' found 'administrative history', 'British history', cultural history' and 'economic history'. Obviously in this thesaurus 'art history' is not a preferred term. There was no reference for what was closest, so I chose the far broader term 'art' in the hope that it would include a term, either as a related or narrower term, that could be used to replace 'art history'.
While the term 'art' was found in the thesaurus it was not even the preferred term, but the synonym of the main term, 'visual arts'. The format of having a table to standardize the record for the term is one that I like. This might just be personal, but I find the even if a section of the table is empty it is easier to see what is there. These descriptors are also very broad categories. Either there is not much in this database about visual art or they have large groupings.
To discover the answer to that question I looked in the database for 'visual arts'. Once again there was no notation in the thesaurus about how many records there were for each term. I found that there were not many records that matched this search. It took an extra step beyond the results list and the full record to find the descriptors. Once I found them I realized that the indexers did not seem to have a limit in the number of terms ascribed to each record. There were over 60 terms for the record, covering anything that could possibly be addressed in the study.
The list of keywords do link back to the thesaurus so that you can see how the term you are looking for is related to the other terms. I tried to follow 'National Identity' to the thesaurus and found the first scope note I had seen.
One of the main differences between the two thesauri are their use of relationships. In the BHI thesauri there are not as many relationships as the UKDA has. To check this theory I went back to the BHI and looked up 'national identity'. Where the UKDA entry has 10 relationships the BHI has 5. I am not claiming that this is representative of the entire thesauri, but it is common in the terms that I compared.