Managing Serials with Evergreen at #EVG10


Stephen Putt and Daniel Wells of Hekman Library at Calvin College presented on their work for serials management in Evergreen. They point out that for many libraries, serials is still a fairly major chunk of technical services, with lots of throughput, many repetitive tasks, and a lot of importance placed on fine-grained details. Any serials workflow needs to plan for predictability, but allow for exceptions, which are frequent. Their work with Evergreen aims to internalize and automate as much of the serials process as possible, so workers can focus on getting the issues processed and out to patrons.

They describe three levels of data in a serials model:

Evergreen Serials Model, taken from http://open-ils.org/dokuwiki

  1. Metadata: the prediction patterns and holdings ranges in MFHD
  2. Physical: where the materials actually are
  3. Business: acquisitions data

In the diagram, the tables in green are the serials tables, while the other colors represent parts of the core Evergreen database. Links to Acquisitions are not shown, as this is concurrent development which will hopefully be merged together soon.

Most of the work so far as centered around the Metadata level. The MARC Format for Holdings Data builds prediction patterns in as part of the standard, with field 853 including the caption and pattern. There is generally one per bibliographic record, at least to start. The MARC 863 field holds instances of the pattern, either individual issues or ranges thereof. Naturally there would be many of these per record. Other pairings exist for supplements and indexes, and non-standard holdings can be captured textually in field 866.

Each ‘piece’ of a serial has it’s own entry in the issuance table. This table is made up of the expected date of the issue, the actual received date, label info, claiming info, and something called ‘shelving units’. Shelving units are generic containers for issues, linking multiple issues to a single call number. This can be configured to match the libraries’ practices: If you don’t circulate serials, do it however you like. If you circulate current serials, assign different units for each current issue, and one or more for older issues. If you circulate bound volumes, invert that. There are auto-population tools, Auto Per Issue and Auto Per Volume, to help streamline the workflow.

Again, as mentioned above, business information has been put on hold, so that it can be integrated into the work being done with Acquisitions.

The typical workflow for serials in Evergreen would be like thus:

  1. Create the MFHD for the bibliographic record, with the prediction pattern.
  2. Use pattern to generate the issuances.
  3. Receive an issue as it comes in
  4. GO TO 2

Exceptions to this would include placing claims, dealing with prediction pattern exceptions (like canceled issues) and the manual editing/management of shelving units; for example, some binders limit bound volumes to a certain thickness, forcing libraries to split up a volume into two items.

The presentation was followed by a live demo, and some Q&A. One ‘missing’ function so far is any kind of ‘unreceiving’ of issues, though audience members noted that the MFHD could be edited directly in that case. It was also noted that barcodes can be added on receipt of the issue, or auto-added if desired.

Dan Wells wants to point out that he is not a Serial Killer (despite the original intended name for this presentation).

[Originally posted by Ian Walls]

Leave a comment

Your email address will not be published. Required fields are marked *

Are you human? * Time limit is exhausted. Please reload CAPTCHA.