Troubleshooting Hints: Vega Discover Rollups

Vega Discover work level rollups for textual and video materials utilize various BibFrame metadata fields to combine bib records (called instances in Bibframe) to rollup into a combined work record representing all formats and editions of a particular work.  Metadata that does not conform with rollup logic rules can cause titles that are technically the same work to not rollup properly.  Usually correcting small cataloging errors or making format changes to bib records will cause them to rollup correctly with a work.  The information in this help document will assist you in troubleshooting metadata discrepancies in your bibliographic records that are affecting Vega Discover rollups.  

This solution was updated on February 14th, 2024 after changes were made to rollup logic for titles and authors.

I have a title that is not rolling up 

This section of the troubleshooting guide will walk you through metadata corrections you can make to your bibliographic records when you are seeing issues with the rollup of same title records into work level rollups in Vega.

Metadata we will be reviewing for problems

1. Title
2. Author or Publisher
3. Language 
4. Leader/06 Record Type
5. Diacritics appearing in title or author fields

NOTE: There are slight variations in the rules used for the rollup of textual content (books, ebooks, audiobooks, eaudiobooks) and visual content (DVD, Blu-Ray, Streaming Video, VHS).  Please see the book and video icons on each troubleshooting step to see which type of rollup they apply to. 

Step 1. Check the title metadata in the record that is not rolling up, vs an editon that is rolling up. 

1. Do the titles match exactly?

In order for editions to rollup into a work level rollup the titles must have an exact match.  These are the rules for title matching that you should check your title data against: 

1. 245 match on $abnp 
2. mainTitle and subTitle, partNumber and partName are all used for matching if present and must exist in all editions record for them to rollup.  The order of partNumber and partName subfields does not matter.  Keep in mind that each subfield is separately matched so a title with subtitle information must match both the values in $a and $b exactly to rollup.
3. Non-filing characters specified by the incoming 245 title tag are ignored (the number of leading characters to ignore is indicated by the value in the second indicator for the 245)
4. mainTitle string matching criteria:
   1. Title match is case-insensitive and ignores trailing commas, semicolons, periods, and colons, and leading and trailing spaces
   2. Trailing text in parentheses of mainTitle is ignored (e.g., "The Elephant (the movie)" matches "The elephant")
5. subTitle string matching criteria: The following punctuation is ignored: .,;[] as well as leading and trailing spaces.
   1. Subtitles of "a novel" or "a thriller" (case insensitive exact match, ignoring leading and trailing spaces and trailing punctuation) are ignored.
   2. Any text found within [] brackets and the brackets themselves are ignored.

Matching Fixes

Correct any metadata that is causing a mismatch based on the rules above.  After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today. 

Step 2. Depending on the type of rollup, check the Author/Publisher metadata in the record that is not rolling up, vs an editon that is rolling up in the format tab that you expect. 

1. (Textual Works) Which Author is being used for each edition?

Vega BibFrame chooses the primary author based on the following rules: 

1. The first 1xx field is chosen as the primary author.  All 1xx roles are considered valid for selection as primaryAuthor.
2. IF there is no $e role value Vega assumes that the role is 'author'.   

2. (Textual Works) Do the author names match exactly?

In order for editions to rollup into a work level rollup the authors must have an exact match.  These are the rules for author name matching that you should check your title data against: 

1. Authors metadata used for matching is $ae of 100, 110, and 111 tags.
2. Author string matches are case-insensitive and ignore trailing spaces, commas, semicolons, and periods

Author Matching Fixes

Correct any metadata discrepancies between mismatched author fields.  Note that the order of data in the $a must match exactly except for the ignored punctuation mentioned above.  $d dates are ignored for authors.  After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today. 

1. (Visual Material Works) Which Publisher(s) are being used for each edition?

Vega BibFrame selects publisher data for rollup matching based on the following rules: 

1. All 260$b and 264 (2nd indicator must be 1) $b are considered for matching.  If there are multiple publishers found all must match exactly across all editions for a rollup to occur.  
2. Author fields are ignored for moving image (video) materials

2. (Visual Material Works) Do the Publisher(s) match exactly?

Vega BibFrame matches all selected publishers based on the following rules: 

1. Publisher data matching is case insensitive and ignores trailing punctuation ;:.,/, and leading and trailing white spaces. 
2. The list of publishers is sorted alphabetically, and then strings are concatenated using "|" as delimiter before matching is applied

Step 3. Check the Language of the editions

Language Matching Problems

Vega matches the language of editions for work level rollups so that works presented in different languages do not rollup.  Vega determines the language of an edition by checking each of the language fields below in the order given.  If Vega finds a language value in any of the fields it uses the value it found and will not check the other field values from the rest of the list: 

1. 008 language code
   1. invalid/unusable codes are 'und', 'mul', 'zxx', '|||', '###'
2. 041 $a (if 041 is repeated, all languages in the editon must match for a rollup to occur)
3. 130/240 $l (Vega attempts to interpret the text in this field and map it to a language code)
4. If none of the above are present Vega assumes the language of the material matches the default language of Vega Discover set in the Localization settings of the Vega Staff Admin.

Language Fix

Correct any missing or erroneous 008 language codes or other language fields.  After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today. 

Step 4.  Is the edition that is not rolling up an Audiobook or eAudiobook?

Record Type Error for Audiobooks

In order for audiobooks to rollup with other textual works the Record Type (LDR06) must be set to non-musical audio (i).  

Record Type Fix

Change all audiobooks and eaudiobooks LDR06 values to i for non-musical audio to avoid this problem.  After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today. 

Step 5.  Does the title or author include a diacritic character?

Diacritic Problems

In order for editions to rollup into a work there must be an exact match of the characters in the title, author, and language fields.  Sometimes the character used in the metadata record looks the same in the catalog but is actually a different diacritic character.  

Diacritic Fix

Copy the word or name with the diacritic from one of the the rolled up editions to the edition that will not rollup, replacing the diacritic in that record.  After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today. 

Troubleshooting Other Rollup Issues

This section of the troubleshooting guide will walk you through troubleshooting steps for other specific issues you may encounter with rollups in Vega that can be solved by updating your metadata. 

Metadata and settings we will be reviewing for problems

1. Material Type
2. Leader/06 Record Type
3. Record Suppression Status in ILS
4. Vega Staff UI Record Extraction Status
5. Author or Publisher

Title is Appearing in the Wrong Format Group

1. Check the Material Type set in the Bib record.  Is it configured as expected?
2. Check the LDR06 Record Type code.  Is it configured as expected?
   1. An instance will be classified as type = print if LDR06 = a & LRD07 = a, b, i, m, s
   2. An instance will be classified as type = electronic if one of the following conditions is true:
      • LDR06 (record type) = m, indicating a computer file
      • 007 (00) = c, indicating a computer file or electronic resource
   3. An instance will be classified as a type=movingimage if LDR06 = g

After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today. 

Expected Bib Edition Cannot Be Found in Vega Discover

1. Check that the record is not suppressed in the ILS and unsuppress if needed
2. Use the ISBN or recordID lookup urls to search for the record in Vega Discover: 
   1. https://XXXX.YY.iiivega.com/search/card?isbn=ZZZZZZZZZZZZZZZ
   2. https://XXXX.YY.iiivega.com/search/card?recordId=ZZZZZZZZZZZZZZZ
   3. If you receive a 404 the record is not indexed in Vega Discover: Check the Record Extraction status in the Staff Admin for errors and report to Support
   4. If you find an unexpected record, find the record in the ILS and check for and correct any problems with:
      • Uniform Titles
      • Authors that do not meet the criteria for Primary Author (1xx/7xx has $e that is not aut, author, edt, editor) which can cause the title to rollup with an identical title that has no author. 
      • Missing or Generic Publisher Data which can cause the title to rollup with an identical title that has no publisher or generic publisher data. 

After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today. 

 Unexpected Bib Included in Work Level Rollup

Many libraries experience problems with single word titles that have no primary author field to help differentiate them from other works with the same one word title. This is often the case for Travel Guides and some Children's books. In these cases you may have to add information to the title or subtitle field of the instance records to differentiate them and unroll them from a larger work. 

1. Check the title for any metadata discrepancies.
2. If the bib is for a single word or simple title that and has no author, consider adding distinguishing data to the title field to unroll it from the larger work. For Example: A travel guide about Thailand that is simply titled "Thailand" and has no author can be updated in the following way to differentiate the instance from other works with the same title and no author: 
   
   1. Include the Guide Series in the title: $a Thailand $b (Lonely Planet)
   2. Include the publication year in the title: $a Thailand $b (Lonely Planet 2023)
   3. Give the guide series an author designation:
      1. Title: Thailand
      2. 110 Author: Lonely Planet Travel Guides
   4. Some libraries choose to have a single bib record for guides with yearly volumes for each reprint of the guide and use the item volume details to allow the patron to select a particular year's guide.
3. For Movies check the Publisher data.  Some movies of the same title may be published by the same studio and will rollup when they should not.  The only option to fix this is to add metadata to the title or publisher field to differentiate the works and cause them to unroll. 
4. Does the work have no author?  Check for records missing a 1xx field, which can cause the title to rollup with an identical title that has no author.  This is common with children's books with titles like "Snow" and other common words. 

After editing MARC data, save the record in the ILS, and run a MARC record extraction in Vega to reingest records that were updated today.