Meeting Minutes for 10/23/2019

Marmot Union Cataloging Committee
Wednesday, October 23, 2019
Minutes in blue

Announcements

• Montrose is live.
• The Montrose records are in the system.  Lloyd is working on adding some of the OCLC numbers they had in their Horizon system where our records don’t have them, without creating more of the double 001 fields that were created by the Bemis load.  There will only be about 1,500 of them.
• Another Marmot basic cataloging training will be held on the 24th at 1.  Contact Lloyd to sign up.
• Lloyd is going to the Charleston Conference in 2 weeks.

Completed action items

Discussion Topics

• Deduping, cataloging, and grouping Wonderbooks/VOX books
  • This is a new format.  They are a paper book with something like a playaway reader attached to the book, so the book reads itself to you.  Currently Wonderbooks and VOX books each have separate format facets in Pika. Lloyd asks if they are similar enough that they could be in the same Pika format.
  • Nina says they are picture books with a triangular reader unit attached in the corner of the book.  She says there is also music and sound in addition to just reading the book. It is a problem that they have the same ISBN as the regular book.  She asks people to be careful to not combine these with regular books even though they share ISBNs. These will have differences in the 33x fields that catalogers should check for.  They are called audio enabled book. Sometimes that will be in the 300.
  • Betsey says they have both of them and she does not think they are different enough to warrant separate format facets in Pika.
  • Ashley asks, if they share a format facet, what should it be called?
  • MARC records call them “audio enabled book”, but that may not mean much to patrons.
  • Nina says they call them “read-alongs”
  • We could call them “read-along books”
  • Duplicates team should consider terms to cover both Wonderbooks and Vox books and any other similar materials. << ACTION ITEM
  • Since these share an ISBN with the base book there are likely to be problems with record loading that matches on ISBN.
  • Lloyd points out that title verification on the loader won’t help because they will share the same title.
  • Jamie asks what will happen when these are loaded from a vendor?
  • Shelley says they will usually match and attach an item to a record for the base book.  If there is already a record for both, then it will insert a third.  
  • People should be aware when these are in a shipment, and be careful.  They would have to be cleaned up. Either deduping extra inserted copies or downloading new records.
  • Jamie suggests the idea of making a consistent change to the ISBNs for Wonderbooks, like a suffix.  That way they would overlay each other and not the base book. This would only work if we all used the same suffix.
  • Jamie asks if the Marquis plugin could help with this?
  • Lloyd says that Marquis could be programed to deal with this.  The Marquis process does not attach based on 001 or ISBN. It picks the record it wants to overlay and matches on the Sierra record number to overlay a specific one.  It is plausible that it could avoid this completely, but it would require extra programing. Also only a few of us are using it, but it might help, but it would be extra work for Telluride.
  • Shelley asks if anyone knows how long they last.
  • Betsey says they have had them for 6 months.  They seem very durable.
  • Nina says they have USB chargers.  They make sure they are recharged before they circulate again. 
  • Betsey says they will do 150 replays before they need to be recharged, and they have a warning light that they need charging.
• Update on Pika grouping project
  • Pascal reports on the project.  They are trying new grouping ideas on a new test server.  The old test server is 2. The grouping test server will be 3.  So if the address for your regular test server is mesa2.marmot.org, then the address for your grouping test server will be mesa3.marmot.org.  The grouping test server leaves out all of the manual grouping so we can see what the system is doing on its own. Pascal expects to ask catalogers to examine how this is working as he makes changes over the next several months.
  • Another issue is what Pascal calls a null works.  These have none of the grouping factors, so they all get grouped together in one giant group.  You can find these by doing a blank search on opac.marmot.org. This group will be the first hit on the list and it is labeled “Title not available.”  If you do this search in your scoped Pika, you can find your own records to see what could be done to get them to group. They need a title or author so they find a group.  Pascal suggested that libraries check these to fix them so they will get grouped. We realize that one reason that Pika does not see these titles is that the records are being created with non-MARC titles instead of 245 fields.  Pika does not recognize a non-MARC title. These could be converted to 245 fields, then Pika would recognize them.
  • Jamie points out that the open search is useful when you want to use the facet limiters, so having this weird record come up first for that search is not helpful.
  • We could set a cataloging policy that all records must have a 245 field.
  • We could send out a procedure for everyone to check their records and make sure they have a 245 field in all records.
  • We consider adding this requirement to the cataloging standards document.
  • There is no objection to adding this to the cataloging standards document.  Lloyd will make a proposed addition and bring it to the next meeting. << ACTION ITEM
  • Pascal says that the grouping project is being worked on in the Discovery team as well.  Anyone particularly interested in grouping might also want to attend Discovery meetings.
  • Nina asks Pascal to discuss the status of manual merging and unmerging issues.
    • Pascal says that there are situations where doing the manual merge or unmerge does not cause a new extraction from Sierra, so it doesn’t change in Pika.  Grouping won’t change if there is not a new extraction. He is testing some solutions to this problem on the test server.
    • Ashley points out that it does not always fail on the production server.  It is hit and miss.
  • Nina says there’s a problem with the grouping of the movie “The Lion King.”  The new version is grouping with the old version.
    • Ashley says she has been working on that.  She ungrouped the new version from the old one.  But we still have the problem that once a record is on the list of records not to group, it can’t be regrouped again.  This means if we ungroup the new one from the old one, the various versions of the new one can’t be regrouped with each other again.
    • Nina is not happy that this results in the old version being listed at the top.  
    • Pascal says that the old version, which is still grouped, has more bibs and checkouts, so it gets ranked higher.
    • Nina says that the new date should be more important to ranking.  The old one has been around for years so it will have many more checkouts, but it’s the new one that people want now.
    • Lloyd suggests that it might work better if we ungroup the old versions from the new one rather than ungroup the new ones from the old one.  So the new one stays grouped, and the old versions are split up. We don’t know the effect that would have on ranking. Ashley says she will try this experiment to see how Pika sorts things in that case << ACTION ITEM.  Then we can decide if we want to make recommendations for grouping procedures.
    • Nina points out that the new and old Lion King movies are almost the same running time, so even if we use running time to distinguish movies, these two would still group.
    • Lloyd suggests a mechanism where people could manually boost the ranking of a something to avoid the problem where the old version gets at the top of the list because it has more checkouts.
    • Pascal says that’s possible, but would require programming.
  • Pascal discusses the ongoing grouping project
    • He will start with trying to use language as a grouping factor, so that translations of things would no longer group together. This will make language the forth grouping factor.
    • He wants to create some other new grouping categories including graphic novels and Braille.
    • Then he will start experimenting with different movie grouping ideas.
• Nightly Pika extract, can we stop it now?  What other functions does this still perform?
  • This still does the EDS extract for CMU, Adams State and Fort Lewis.
  • Pika no longer uses this to get records because it is now using an API to get the records from Sierra.
  • Lloyd wants to find out if these records are being used for any other purpose.  If we can figure out how to replace this extract with normal Scheduler extracts, then we can make better use of the 2.2 million record bucket in Create Lists that has been dedicated to the Pika extract.
  • Lloyd has already started doing the monthly Gold Rush extracts with Scheduler.
  • One option is to do separate extracts for each CMU, Adams, Western and FLC.
  • This process also sends records for four different vendors for SD51.
  • Jamie asks if the new API system could be used to get the records out for EDS, so we don’t need Create Lists at all for this.
  • Pascal says that might be possible, but we would have to create a program to do it.  It would not be easy. The current API system does not pull out whole MARC records it pulls them out in pieces and reconstructs them.
  • Lloyd mentions that this explains why he has seen that the MARC records displayed in Staff View in Pika are completely out of field order which makes them very hard to read.
  • Lloyd points out that there are many things we need to extract records for and III makes it very difficult and expensive to get the data out the way we want it.  It would be very valuable to have an extraction tool that we could control.
  • Pascal says that the old Pika extractor is also creating something called Adams State Monographs.  We ask if Mary knows what that is.
    • She does not know what it is. However, if we include Adams in the extract because of EDS, then it would get whatever this is for Adams anyway.
  • Nina asks, what is the downside of continuing to do the current process of extracting the records?
    • Lloyd says the problem is that we have to dedicate a giant Review File for the extract.  File 2 can hold 2.2 million records. It needs to be that large so it can hold our entire collection.  If we don’t use this process, then we can make better use of those records in Create Lists.
    • Pascal points out that this full export requires a special process that III had to write for us.  One time, this process failed when we did an upgrade, because III did not account for our special programming in the upgrade.  This could happen with any future upgrade to Sierra. Using standard Scheduler rather than special programming would eliminate this risk.
  • Lloyd realizes that we might be able to create a file that only included the records that need to go to EDS and SD51.  That file might be small enough that it could be exported by Scheduler. Lloyd will experiment with this possibility. << ACTION ITEM  
  • Even if scheduler can’t export it and we had to use the III black box, but it would allow us to save many files from file 2.
  • A chat asks what is EDS?  
    • EDS is EBSCO Discovery Service.  It allows a combined search of the book catalog and journal articles.
• Weird characters appearing in records
  • These are messed up diacritics that have been appearing in the catalog.  We think this was being caused by the Marquis plugin that some members use for acquisitions.  We are working with Lisa at Mesa and Andy and Telluride to figure it out and we think we know what is wrong.  It should be fixed in the new version of Marquis.
• Discussion of changing subject heading “Illegal Aliens” in the catalog
  • Robin at Bemis has asked if Marmot can change our catalog so that we stop using the term “Illegal Aliens.”
  • The background on this is that several years ago the ALA council passed a resolution asking Library of Congress to change this heading.  LoC decided to split it into “Non-citizens” and “Undocumented Immigration.” The House of Representatives passed a resolution against making any change to the heading.  Although the Senate never took any action, LoC decided to drop the issue and nothing has happened since on the Federal level.
  • Recently, several libraries, including CU Boulder, have taken it upon themselves to change this heading in their catalogs on their own.
  • We could use our authority system to change these headings in Sierra.  How we would do it technically is uncertain. We could probably get Marcive to give us a non-standard authority record and change our headings.  I would guess this is common enough now that they have a procedure in place, but we have to ask.
  • Alternatively, we could use the automatic authority control in Sierra.
  • This would be difficult to manage as a split as LoC was planning on.  We would want it to be a single different heading and make a simple flip.  So we would have to decide on a single replacement term.
  • There are really two different questions here.  There is the technical question of how to carry it out, and then there is the policy question of whether we want to do it at all.  
  • The policy question should be decided by the Marmot Directors.  We could make a proposal and send it to the Directors.
  • Jamie asks if we would use second indicator zero on these.  VuFind has the ability to limit a search to LCSH, and without the zero this heading could be excluded from a search.
  • We could treat it as LCSH, even if it isn’t official.
  • Jamie asks if we should figure out the technicalities before we ask the Directors or the reverse.
  • Shelley says that we should figure out the problems first.  We should just give Directors a simple choice without loose ends.
  • Mary says she would support changing to “Undocumented Immigrants”
  • Nina also supports the change.
  • Jamie supports it as well.
  • Lloyd mentions that “Undocumented Immigrants” does exclude undocumented temporary visitors and suggests we consider “Undocumented Aliens” which would include those people.
  • People think that the term Aliens is confusing, and prefer Immigrants.  We can try that and change it if there’s a problem.
  • Lloyd will write up a proposal for the next UCC meeting << ACTION ITEM
• We discuss changing the UCC meeting to November 20 instead of the day before Thanksgiving when few people will attend.  We also decide to try to change the December meeting a week early as well.

New Action Items

Ongoing Action Items

Next Duplicates Sub-committee meeting: November 13

Next UCC meeting: November 20