Identifying close matches
We are coming across some items which we have already scanned by are not being recognised because the request information doesn't exactly match the biblio of the previous request(s).
This is causing us to rescan items or order CFPs where we don't need to, costing us time and money.
If the vault search could identify other requests from the same publication using ISBN or book title (view all requests for this book doesn't seem to be too reliable) and allow us to preview these scans we could then select the appropriate scan.
Following feedback from users, we identified that the vault was too specific in its matching to existing pdf’s in the vault.
As a starting point in finding matches more readily, we have relaxed the logic in a few ways:
It is no longer looking for exact title matches e.g. Oliver Twist will now match with oliver twist, whereas it would not previously, as it was case sensitive.
It will now match variations of chapter numbers in more ways e.g. Chapter 4, Chapter Four, Ch.4, Ch. Four, ch 4, chapter.
We identified that if a university had already purchased a Copyright Fee-pay copy from the British Library and a subsequent request was made for the exact same book section or article it was referring as Source to be acquired. We have improved this behaviour so BL Fee Pay copies are checked earlier in the worklog so they can be identified as owned.
If this is still an area you would like to see further enhancements made please do raise again.
Ian Long commented
We have noticed that TADC doesn't recognise previous scanned items that have been withdrawn, so this potentially causes the library to rescan items again unnecessarily . We would definitely like this to be improved.
Rob Melocha commented
I agree. ISBN and page number are unique identifiers that would be a good criteria for matching wheras chapter titles or book titles can be interpreted and entered in different ways.