Patron - Service Queue - Nightly Selection

I know that the fact that every book is now available has presented its own challenges along with the opportunities - and I'm hoping that in time different algorithms can be explored when filling the service queue. Right now it appears to be more "science" than "art".

It looks like the software is going systematically by subject through the most recent books, first. The problem is that NLS very often produces many books by the same author with adjacent book numbers, meaning that the same authors are coming up for everyone whether they have an interest or not.

Turnaround was so lovely - a recently returned book was likely one that another patron might want, and it was prioritized to go back out to someone else. What if KLAS were to look at recently requested books or recently checked in books as higher in priority?

I know that there is likely no perfect formula, I've just been getting more complaints as of late with the current crop of books.

Thank you for your consideration!

While the turnaround method sounds enticing, the one main difference is that, at least for us, currently books added to the service queue to create orders aren't a by-product of the most popular books checked out or checked in, which is how turnaround shelve work. They're based on the premise that because you have newer books they come back in and instead of getting shelved in with the entire collection they get shelved in a turnaround area because they are assumed to be wanted more frequently. In a virtual environment there isn't a restriction like this or a dependence on finding a physical item to pull to send out. You have all of the books in one place and any book from the collection can be retrieved easily. So I guess the branch out of the turnaround concept for duplication might be one where there is a way for somehow the most popular or most checked out books, etc, can end up on a patron's service queue to then create an order. The scenario might be where in KLAS it tallies the check outs of books across the collection and flags them in some way. Then by some kind of coding these flagged items are put into a pool of items that could be chosen by patrons who have a subject preference or something else as a code that tells the system that the patron wants a books from this set regularly. This set would change over time as more checkouts happen. This could be useful for down the road if we also put into effect some kind of "voting" mechanism for patrons to pick their favorite books and then others can use that voting to automatically pull in books. People like top 10 lists, which is why we do this each month.

I completely agree it would benefit our patrons to be able to prioritize books that have been requested by other patrons. I would specifically focus on books that have been requested or manually selected by staff, rather than books that have been auto-selected.

One way to accomplish this might be to build auto-select to cross reference between subjects when creating a preference. Instead of a patron adding "Mystery" as a preference, they could have "Mystery AND Bestseller" or "Mystery AND Amish" or "Mystery NOT Suspense". You would then need a subject to cross reference that is (ideally automatically) updated regularly to include only the most requested/staff selected titles from the past X days. I believe this could be done manually through reports and the Bulk Update Subjects tool, but I know I would rather KLAS took care of this automatically.

This also has many side benefits for creating more specific preferences, especially for libraries like my own that aren't able to do a lot of detailed local cataloging. Plus we could create a "perennial favorites" subject to cross reference against including all of the authors that have enduring interest but may not have enough requests at any given moment to hit the Turnaround list (or whatever it would be called). Perhaps this subject could also be automatically generated based on an author representing more than X% of total author preferences among active patrons?

I think it would also help a ton for auto-select to be aware of Series Sequence information. For instance, if a book is selected and it has a series sequence between 2 and 10 (for example, assuming anything 11+ isn't part of a tight, multi-book story line) the patron could instead be assigned the lowest sequence title in that series that isn't marked "has had". If auto-select could also limit to one book per series per subject (when selecting by subject) in each nightly run, that would mitigate the problems that arise when the NLS dumps a chunk of books from a single series.