Controlled Vocabularies Compared With Full Text Search
Lesley Mackenzie-Robb, writing on vantaggio-learn.com compares free text searching and controlled vocabularies.
She makes the point that while Controlled Vocabularies should theoretically provide an enhanced search experience because they offer the user a more focussed and subject oriented range of options to narrow searches, there is subjectivity involved in the selection of the tags and other metadata used to develop a Controlled Vocabulary. As a result, the tags chosen for cataloguing may vary from one individual to another – leading to unpredictable results which might not be considered accurate by the using issuing the search or another cataloguer:
“On the face of it, you would think that using a controlled vocabulary – that is a pre-determined list of terms relevant to a particular domain, which can be hierarchically structured and contain lots of interesting relationships – would the method most likely to lead to accurate search results. After all, the user selects a category from a list to reveal all contents which have been indexed to that term. Ah but….Subject indexing or tagging is the process of constructing a representation of a resource that is being tagged. Savoy (2005), for example, suggests that prescribing a uniform and invariable choice of vocabulary terms actually helps in achieving consistency. In tagging a document or whatever resource, you are making a statement about it. According to Vob (2007), this entails a process of conceptual analysis and translation. It is consequently a subjective process and there is no guarantee of achieving high levels of consistency between human subject indexers. In the simplest terms, the filename that you might give to a document one year may hold absolutely no resonance in your memory when seeking that same document in another year. Human memory is notoriously fickle (Loftus, 2002).” [Read More]
She makes the point that while Controlled Vocabularies should theoretically provide an enhanced search experience because they offer the user a more focussed and subject oriented range of options to narrow searches, there is subjectivity involved in the selection of the tags and other metadata used to develop a Controlled Vocabulary:
“On the face of it, you would think that using a controlled vocabulary – that is a pre-determined list of terms relevant to a particular domain, which can be hierarchically structured and contain lots of interesting relationships – would the method most likely to lead to accurate search results. After all, the user selects a category from a list to reveal all contents which have been indexed to that term. Ah but….
Subject indexing or tagging is the process of constructing a representation of a resource that is being tagged. Savoy (2005), for example, suggests that prescribing a uniform and invariable choice of vocabulary terms actually helps in achieving consistency. In tagging a document or whatever resource, you are making a statement about it. According to Vob (2007), this entails a process of conceptual analysis and translation. It is consequently a subjective process and there is no guarantee of achieving high levels of consistency between human subject indexers. In the simplest terms, the filename that you might give to a document one year may hold absolutely no resonance in your memory when seeking that same document in another year. Human memory is notoriously fickle (Loftus, 2002).” [Read More]
b