After about a month of working with the headings validation tool, I’m ready to start adding a few enhancements to provide some automated headings corrections. The first change to be implemented will be automatic correction of headings where the preferred heading is different from the in-use headings. This will be implemented as an optional element. If this option is selected, the report will continue to note variants are part of the validation report — but when exporting data for further processing — automatically corrected headings will not be included in the record sets for further action.
Additionally — I’ll continue to be looking at ways to improve the speed of the process. While there are some limits to what I can do since this tool relies on a web service (outside of providing an option for users to download the ~10GB worth of LC data locally), there are a few things I can to do continue to ensure that only new items are queried when resolving links.
These changes will be made available on the next update.
–tr
Comments
One response to “Automatic Headings Correction–Validate Headings”
Very interesting! I was wondering whether there are any plans to include the BISAC ($2bisacsh) subject headings…
A common form would be: JUVENILE FICTION / Social Issues / Friendship. <- This is the type of heading format that BISG delivers when they ship the whole file of annually updated headings to its BISG members.
In MarcEdit it looks like: =650 \7$aJUVENILE FICTION / Social Issues / Friendship.$2bisacsh
Although in actual use, many records seem to have variations on that basic form. E.g.:
=650 \7$aJUVENILE FICTION$xSocial Issues$xFriendship.$2bisacsh
=650 \7$aFriendship.$2bisacsh
Our library uses those BISAC headings. I have been using MarcEdit during the past year, to find all heading with 'bisacsh' in my eContent marcs before I import them, and thus allow me to manually adjust individual headings. My authority files are much saner now. But if there is a way to more automatically accomplish this, I would be delighted to hear!