I've noticed a couple of brief eLGAR serial records lately which have been added to the NBD through OSMOSIS. One example is Te Puna #12590582. This record only has a 245 and a 260 $b but it has met the OSMOSIS processing criteria. It lacks a 300, but TMQ are now allowing serials without 300s through.

Perhaps the minimum record standard for OSMOSIS should be higher than this? If it was it may reduce the number of old and very poor records which are being added to the NBD through OSMOSIS.

Peter

Can't believe that this is the minimum level allowed for records to be processed through OSMOSIS.
I think we need to set the standards so there is at least one identifier per record for matching.
There are at least 3 other entries on TePuna that could match this record, but only 1 has an ISSN.

Leonard

One of the biggest headaches is that when a record doesn't match, it automatically creates a new record to attach the holding to, even when the record itself is simply very poor in quality.

Maybe we need to suggest some pre-load vetting system that will weed out the most obvious 'poor' records prior to going through the matching process?

On a completely separate issue, you'll see I'm loading some draft material for the OCLC workstream. It'd be helpful if people can read through this before the meeting tomorrow (yes, I know it's a tight timeframe) or even comment on the relevant pages if something occurs to you while reading.

I've made a little change to the minutes - just for clarity

I thought it may be useful to give my thoughts on the ANSI standard for serial holdings before we meet this afternoon. I don't think we can continue with this as the standard for the NUC serial holdings. Osmosis takes a snapshot of the holdings in each Library's ILS. We are all entering our holdings differently (because each ILS has different requirements and we each have our own local practices). The ANSI standard may have been possible when holdings were being entered manually in the NUC, but the the situation is different now. Another point is that libraries using vendors such as Serials Solutions for ejournal MARC records have to accept the holdings style that the vendor uses. Looking at some other library catalogues I think only LIU, DU and WN are using the ANSI style for their holdings.