Richard SaponWhite Oregon State University May 8 2015 ELUNA 2015 About Data Structure and Maintenance Two Examples Lessons To Be Learned Sharing information 2 Overview The fault dear Brutus is not in our stars ID: 589484
Download Presentation The PPT/PDF document "Know Your Data: A Structured Approach to..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Know Your Data: A Structured Approach to Migration Preparation, Post-Migration Clean-up, and Ongoing Metadata Maintenance
Richard Sapon-White
Oregon State University
May 8, 2015
ELUNA 2015Slide2
About Data, Structure, and MaintenanceTwo ExamplesLessons (To Be) Learned
Sharing information
2
OverviewSlide3
The fault, dear Brutus, is not in our stars,But in our data,
that we are understaffed.
3
Shakespeare on Integrated Library SystemsSlide4
“Structured information about an information resource of any media type or format”
– Priscilla Caplan, Metadata Fundamentals for All Librarians. Chicago: ALA, 2003.
Structure on macro level
Bibliographic records, item records, holding records
Structure on micro level
Individual MARC fields
and subfieldsItem record elements
4
What is (meta)data?Slide5
Content rule changesAACR2 RDA
Revision of authority records
Serial title changes
5
Metadata is DynamicSlide6
Status changes: on shelf, checked outMissing or withdrawnPut on reserve
URL changes
Insert notes:
missing pagesmissing issuesautographed copiesEt cetera!
6
Metadata is Dynamic (cont.)Slide7
Bibliographic, holdings, and item data
Fix bibliographic records that do not have a 245/title
Clean up invalid dates.
Clean up chronology and enumeration (item/issue description) for serials as Alma uses these sequences in issue-level requests.
Identify links pointing to existing source systems that are migrated, so that they can be removed or updated, if necessary.
Identify and delete duplicate Bibs.
Consolidate multiple holding records for identical item copies for the same location/call number.
Assign a value for empty collection/location for items/holdings, if applicable.
Statuses and codes
Standardize statuses and codes to remove extraneous entries and consolidate them into fewer options. Review and reduce purchase order statuses, if this is applicable. Review non-standard MARC fields (alphanumeric, for instance), revising records as appropriate. Note that migration allows mapping to local 9XX fields.
Review and consolidate material types for items (for example, BK, Book, Monograph)
Old data
Clean up and/or purge patron records that are inactive or expired and have no outstanding items, fines, fees, or other transactions.
Clean up and/or purge patron records with fines, fees, or other transactions that are older than a certain date and/or less than a certain amount.
Clean up and/or remove purchase orders that are old drafts or tests.
Clean up and/or remove invoices that are old drafts or tests.
Clean up standing orders and subscriptions that are old, should have been closed, or require investigation when active, but have been inactive for years.
E-resources and package resources
7
Getting Ready for Alma Implementation, Appendix B
Optional Data PreparationsSlide8
ILS consists of recordsEach record has component parts (fields, elements, etc.)Each field/element may have subfields, etc.
8
The Structured ApproachSlide9
Consider each component field and subfieldRecord in a central location/document the definition of the field/subfield and what data is acceptable for that field/subfield
Less concerned with content defined by national standards (e.g., RDA/AACR2-specified content, MFHD)
More concerned with locally-applied data and tracking of records
9
The Structured Approach to MaintenanceSlide10
Millennium is heavy with codesSuppress/T field to identify bib records to suppress
10
Case #1: Suppress/T codes
= not suppressed
c = SUPPRESS ORD (10)
r = ACQUISITIONS (382)
p = MARCIVE SERIAL (169)
u = MARCIVE MAPS (0)
i
= BNA SERIAL LOAD (4480)
l
= DON’T USE-ORB (6)
s = NOT ORBIS/SUPP (4479)
d = DELETE CODE (16)
a = APPROVALS (46)
w = WITHDRAWN (1171)
t = MARCIVE MONOG (10,662)
v = MARCIVE MONTHL (136,758)m = CONVERN ON FLY (32)n = SUPP LOC & ORB (1069)z = NOT ORBIS/DISP (28,016)Slide11
No one remembered what some codes meant.
8 records had a blank (not “-”). These were “ghost” codes that Millennium retained, but could not be used in a search.
Elsewhere in Millennium, table indicated which codes meant suppress and which were just markers.
11
Case #1: Pre-migration
Problem (cont.)Slide12
Anything other than “-” migrated as suppressedNone of the codes migrated to AlmaPost-migration clean-up
All
Marcive
records (p, t, v) had 086 and were unsuppressed in a batchRecords marked for deletion were deletedSome had other characteristics that were used to identify them
12
Case #1
: What Happened at MigrationSlide13
For every code or field, document:Its definitionThis field is used to identify whether the bib record is suppressed or unsuppressed
In some cases the field is used as a marker and does not reflect record suppression
The meaning of values
a = approvals (suppressed so that users don’t request materials not yet cataloged)
m = conversion on the fly (not suppressed so users know we own the title, but status indicates unavailable
13
Case #1: Lessons LearnedSlide14
14
Case #1: Lessons Learned
Code
Definition
# of records
Notes
a
Suppressed
520
Use
only for checking in issues
b
Not
suppressed
1,234,567
c
Marked for deletion (suppressed)
5
Delete quarterly; see dept.
manual for details
d
Record load from XYZ,
inc.
954
Records are
not suppressed; delete after lease expires 6/31/2015Slide15
Store your documentation in a form that is easily accessed by staff:Excel spreadsheet
Word document
Department manual in paper or online
Wiki
15
Case #1: Lessons Learned (cont.)Slide16
Has considerably less coding than MillenniumHas many free-text boxes in inventory (especially notes: internal, statistics, internal description)
Let’s look at an example!
16
But Alma…Slide17
In Millennium: Over 5 years, had loaded ~70,000 bib records from EBL via
contract with OSU
001
EBL239893When purchase triggered, DDA record overlaid with one from OCLC
17
Case #2:DDA RecordsSlide18
At same time, OSU participated in EBL DDA program via Orbis Cascade Alliance:
740
Orbis
Cascade Alliance EBL DDA unpurchased title.When purchase triggered:740 Orbis
Cascade Alliance EBL DDA purchased title
.
Post-migration: Manage DDA via CZ
Delete OCA DDA records prior to migration
18
Case #2: DDA RecordsSlide19
Some OSU DDA records matched with NZ records and were overlaid001 EBL number replaced with OCLC numberSome OSU DDA records unmatched and retained 001 EBL number
Purchased OSU DDA records unmarked!
Ability to do monthly loads of OSU DDA records was problematic
19
Case #2: What Happened at Migration?Slide20
Purchased titles identified with Internal Description note in portfolio (~2000 records)EBL purchaseCreated set of IZ-only records for deletion
Marked unmarked records in NZ with “OSU EBL DDA to be deleted” in Internal Description note
After everything deleted, reload entire set of records for a fresh start
20
Case #2: Problem resolution Slide21
Import profiles – use the Profile Description box!Mark sets of records so they can be isolated and manipulated later
Delete
DDA records pre- & reload post-migration
Create a log of record loads and their distinguishing characteristics (history in Alma retained for 1 year only)
21
Case #2: Lessons LearnedSlide22
If pre-migration, look at every unit of (local) data and…Record its definition/description
Determine if it needs to be edited or deleted
Can editing or deleting be completed pre-migration?
If not, will the data migrate so it can be edited or deleted later?
22
Get to know the data you haveSlide23
Whether pre- or post-migration: Create a central location to record information about your data
Make sure everyone in your unit/library knows where to find the info
Review it periodically for updating!
23
Best (?) PracticesSlide24
Richard Sapon-WhiteHead, Cataloging and Metadata Services Unit
Oregon State University
richard.sapon-white@oregonstate.edu
24
Thank you!