So many records, so little time
Download as PPTX, PDF1 like359 views
The University of Limerick Library outsourced the cataloguing cleanup of over 300,000 records to improve consistency and prepare for uploading records to OCLC. The cleanup standardized formatting, applied authority control and Resource Description and Access enrichment according to the supplier Backstage's program. It took 3 months to load back the cleaned records 5,000 at a time and addressed challenges like location name changes. The benefits included records ready for OCLC, consistent formatting, and improved discovery capabilities.
So many records, so little time
- 1. So many records, so little time … UL Library’s experience of outsourcing catalogue clean-up Justine Bennett Librarian, Cataloguing & Metadata
- 3. • We wanted to upload our records to OCLC • Authority control • Resource Description & Access (RDA) enrichment • Satisfaction from cleaning Why do a ‘clean-up’?
- 4. • National Library of Ireland recommendation • Customisation options provided • Resource Description & Access (RDA) programme offered • Relationship existed between UL & Backstage Why this supplier?
- 5. • Create our online profile • Determine our RDA requirements http://ac.bslw.com/mars/guide/ • Decide reports schedule and requirements Agreeing terms for the project
- 6. • Sooooo many reports • Loading back 300,000 records – Did 5,000 a day – Took us 3 months Implementation
- 7. Our challenges 1. Scarriff, Co. Clare becomes Colorado . . . 2. Provenance field (561) duplicating on OPAC
- 8. Improvements and benefits • Records ready for OCLC upload • Consistent formatting and updated catalogue records • Display of correct headings in user searches • Elimination of manual, cumbersome processes • Future discovery capabilities based on RDA fields
- 10. Justine Bennett Librarian, Cataloguing & Metadata Acknowledgments Cutlery ‘before’ http://tinyurl.com/omas3ta Cutlery ‘after’ http://tinyurl.com/nqpa64x With special thanks for their help on this project - Gerardine Ahern, Judy Carmody and Caleb Derven.
Editor's Notes
- #3: Cleaning the catalogue is like cleaning the cutlery.
- #4: OCLC reported back to us after we had loaded a sample and explained that there were too many errors and inaccuracies. Most common errors - 008 field were the date & language were missing or they appeared as invalid. We wanted to merge and correct any unauthorised headings. Authors names needed to be tidied from multiple entries to one main entry The authority control process standardises name, subject, series, and uniform title headings. RDA records were already being imported into the catalogue and we wanted some consistency. If you saw cutlery like this you’d have to wash it to use it!
- #6: We scheduled a conference call with Backstage and filled in an online profile. There was a hover button explaining each option which was helpful. To be really honest with you we chose default for quite a few of the options as we were unsure how it would effect our records. In the RDA profile again there was a template to work with. I spent more time with these options 1) because I was not on the phone to them and could do it at my own pace and 2) RDA at the time was so new to me that in reading thru the options I was understanding what RDA involved at record level! The RDA link is to narrative on statistics of what clients have chosen to have enriched since 2013 when RDA began. Reports – the vendor had offered weekly reports of work they had carried out but they also suggested that other clients had been overwhelmed by these very detailed too frequent reports so we decided to go with a monthly and bi-annual authority updates reporting option.
- #7: So these reports would come rolling in and I would be checking through them and think I was doing a fab job but there were challenges along the way. I will talk you through some of them and how we in UL solved them. First sample in January, reports in June checking on my own for few weeks. Checked records on our MIS asked Special Collections to check their records on MIS. Now this took a lot longer than it did for me just to say that sentence. There was so much to check (plus there was bi-weekly status reports of our project) so I asked our rep @ BSLW what would be the top 5 main areas worthwhile checking. By Summer we were confident to load the records back into our LMS. My cataloguing colleagues and I loaded 5000 a day with help from Capita.
- #8: Co. becomes Colorado. In extending the abbreviations of Co. it had now become Colo short for Colorado and not Co. I explained this to BSLW and they were able to run a script to find records and then send us back the fixed bibliographic records. All 2581 of them. Back on track colleagues in Special Collections noticed the provenance field 561 was repeating itself in the OPAC. After some delving/blaming it turned out to be a local issue where this field protects data at all costs. These were fixed using a tool called MarcEdit. The 300 000 records were now cleaned, shined and ready for loading into OCLC for collection analysis.
- #9: 2. As Helen Williams of LSE says it is only possible to retrieve relevant records from a catalogue if there a degree of standardisation in the way the catalogue is organised.