Why do the number of records in my query collection differ from the number in MarcEdit when I upload the file?
Applies to
- WorldShare Collection Manager
Answer
Collection Manager indexes are rarely 100% correct for these reasons:
- Merges that haven't been reflected in the index yet could cause higher result counts. For example, OCN 1 merged into 2. Both could be in the result, but only record 2 actually exists (2 would appear in the search but only 1 would appear in the file).
- Sometimes WorldCat records are just deleted without being merged into anything, and these are often still included in query collection results.
- Sometimes the numbers returned by a query were never valid OCLC numbers at all or at least point at records in Worldcat HBase that either do not have the required dataSource="xwc" column or are just blank in the "document" column where the actual XML record is supposed to be.
- Sometimes records cannot be crosswalked into the desired format (marc21, marc xml, etc) for issues such as too much data in a field / subfield. This can be caused by the WorldCat record itself or by the user's customizations in the institution profile.
Additionally, if you have applied MARC Delivery Filters, these are not accounted for in the estimated number returned by the query. This estimate is determined only by the WorldCat Selection Criteria, and the filters are applied afterward when the system is outputting records.