Tuesday, 3 May 2016

Database cleanup work (Technical)

Alert – this post contains some very important information about improved user experience – but also some technical stuff that not all of us will fully understand.  (Much of it has been written by Di Cranwell our cataloguing guru!)

PLS has engaged the services of the Database experts at SirsiDynix to complete a different to usual Database clean-up  This process will be followed by a load of Library of Congress Subject, Author and Series headings. These processes were originally scheduled for May, but given our recent network instability we have postponed this until the next available “window” in the LMS schedule to undertake this work.  It has therefore been rescheduled for August.

All Database clean up activity has a primary objective of improving the user experience in using the system – whether the users are library staff or customers. So this is something that is uppermost in the minds of many who use the system. We are aware that the database continues to contain records from various pre-One Card local cataloguing conventions, some of which impact on the user experience. For instance different libraries had their own conventions for the use of GMDs (general media designators) in Titles. Some used [Music CD], while others used [music – cd], and other errors crept in such as [musci CD]. Because these diverse terms were in titles the same items with different GMDs could not be matched by our previous automated database de-duplication processes. However our new database experts have found a way to map and change all erroneous GMDs to standard ones. Oh – and BTW the standard for music CDs is [music CD]. As soon as we get correct GMDs in records we can merge those which are for the same work.

And this same logic can be applied to other AV materials where the Material Type in the record is clear. For example if the Material Type of the record is “DVD” and the existing GMD is [videorecording] this will be updated to [DVD]. GMDs that are no longer used e.g. [book] and [text], will be stripped from the title. Once this work has been completed, we will reissue the list of One Card approved GMDs by updating the LMS Ops Guide.  There will be an expectation that libraries use these in new Bib records ensuring consistency in the future.

A clean-up of the MARC Tags in the Bibliographic records will also take place.  Tags that are considered “junk” Tags that have been unintentionally loaded into the database will be removed along with obsolete Tags.  This part of the clean up will strip over 300,000 lines of data from our records; leading to a cleaner and faster database. 

Some Tags will be updated to the latest standard, for example obsolete series fields 400, 410, 411, and 440 will be upgraded to field 490/8XX pairs. There are also many local Tags which were used during data loads, either with the One Card implementation or as used by previous Library Management Systems, that will be deleted. RDA tags will be added where possible and as RDA Cataloguing rules prohibit the use of abbreviations (unless they form part of an actual word in a Title/Author) these will be updated to the full version of the word if identifiable e.g. Dept. becomes Department.  SirsiDynix has also offered to add a 007 Tag using the Item Type of the item as the source of information. Therefore if the Item Type is AB [audio book CD], the 007 Tag will be added and this will update the icon on Enterprise from a “book” icon to an “audio disc” icon.

All of this will mean much tidier Bibliographic records, reducing the indexing reports duration and increasing search speed for library staff and search results appearance for our Enterprise customers.

The second part of the Authority Processing Service will be a load of Library of Congress Subject, Author and Series headings including the Children’s headings.  These will be added in addition to the existing Libraries Australia headings loaded in 2014 so that no local headings will be lost.  A report will then run to match the Tags in the Bibliographic records to the correct form of the heading and update the Tag.  Again, this will clean up and enrich our records.

There has been much discussion with SirsiDynix regarding our local headings for SCISS, Torrens Toy Library and Local History.  All of these headings will be retained.  Where the Bib record indicates the Audience level to be “Juvenile”, a subdivision “v” of “Juvenile literature” or “Juvenile fiction” will be added.  If the Form subdivision is incorrect, e.g. Fiction in an “x” subfield instead of the correct “v”, these will be updated as well.

The LMS Collection & Cataloguing Group have been eagerly awaiting the Authority Processing Service since it was first presented at a meeting in August 2015 and the results will be appreciated by all Network staff and customers.

I would like to thank the LMS User Group members for their support, along with Jo Cooper as the Chair of the Collections & Cataloguing Group and the members of the sub-group who have been working through the finer details to make this all happen; Alice Mariano (Holdfast), Chris Kennedy (PLS & Holdfast), Angela Jones (Salisbury), Peter Thomas (Mitcham), Cathy O’Brien (Campbelltown), Jane Murphy (PLS) and Di Cranwell (PLS).

And I should add that once all of this work is complete I believe that our database will be sufficiently clean to add all of our holdings to Libraries Australia.  But what that means will be the subject of another post some time in the future.

1 comment:

  1. This all sounds great... It will be so much easier for all of us to have single records for items... Keep up the awesome work PLS team!!