Search Type Considerations
Choosing the types of searches to allow a user to perform requires a
balance between providing a simple interface and allowing specialized
access to the data. It is worth remembering that most users do not
understand or take advantage of advanced search features. That said,
it is also desireable to provide access to depth of data if you have
it.
Things to consider when specifying the types of searches needed:
- Is your source data sufficiently differentiated to support fielded
searching?
- How many records will the database contain? If the database is
relatively small and/or the data in the fields relatively unique, then
field-specific searches might be overkill.
- consider whether a field-specific search would provide results
similar to the same search done as a general keyword search.
(e.g., if typical search strings do not appear anywhere other than
in that field, then a general keyword search would suffice.)
- consider whether the number or records is so small or data is
repeated so infrequently that general keyword search result sets
would be small enough to scan easily.
- What types of information might specialists in the field want to
search by.
- if specialists know a particular data string or are interested
in a particular data type, it might be useful to provide fielded
searching to support targeted retrieval.
- if specialists might want to retrieve specific subsets of
information (e.g., only items published in a particular location
or after a particular date, etc.) then field specific searching or
limiting might be needed.