Wednesday 23 April 2014

De-duplication continues to deliver benefits

As anyone using the system will know, the issue of duplicate bibliographic records continues to be one of our biggest hurdles to efficiency, in terms of searching for items in specific libraries, managing "holds," the number of items in transit, cataloguing to the correct bib records and other areas.  Some libraries are performing manual "transferring" - i.e. resolving these duplicate records, which is to the benefit of all who use the system.

To attempt to speed up this process we engaged SirsiDynix to run a de-duplication script across the database over Easter, with the intent to merge as many bibliographic records as we could find exact match points for. The script has managed to merge over 20,000 records, which is a great outcome.  

The process also identified a further 1,300 records which could be merged, however they have items with the status of "in transit" which stops the merge happening.  Therefore the script will continue to be run a number of times over the next week or so with the intent of resolving these items as their status changes from "in transit".

We will continue to run de-duplication scripts from time to time, however there are a significant number of records on the system which cannot be resolved using automated scripting.  Therefore we will need to use a concerted approach of staff fixing these intractable records.

PLS will be working with the LMS User Group to identify an equitable way of assigning this transferring work, with the aim of having a much cleaner database in the coming months.  We will keep you posted as to how this progresses.



No comments:

Post a Comment