This submission as #6795 was 2nd ranked in the Alma 2020 NERS round, to be delivered in 2021, with high level conceptual design explained in Itai's comment of December 20, 2020.

(shameless plug)

See also https://ideas.exlibrisgroup.com/forums/308173-alma/suggestions/43847271-allow-to-define-several-local-search-indexes-incl

"Allow to define several local search indexes, including standard MARC fields" - a more modest suggestion, that hopefully can be more technically feasible and still provide a large chunk of the missing functionality

Hello,
Thanks for the update.
Unfortunately I think the proposed wrkflow is far too complicated for the average librarian. It might be OK for a system librarian or a metadata specialist, but nor for most of our staff. They are not used too using sets and indication rules.
Something like the mockup that Stacy van Groll attached to her comment looks far more usable to me.
Best regards.

As part of the design process, the Alma team would like to share with you the conceptual design. The flow is described at a high level. Details are discussed with Stacey van Groll (as the representative of NERS for this request #6795). Additional review of details will be done with Alam UX Focus Group.

The basic flow:
1. A user will conduct an advance search.
2. In case a large number of results will be returned the user will be able to further filter the results.
3. The user will choose the action "save and filter query" (currently support save query)
4. In the "save and filter" page the user will define the logical set details as well as choose an indication rule which will have all the additional conditions in level of field and subfield.
5. In case the relevant indication rule doesn't exist the user will be able to create it via Metadata Editor as a form-based indication rule builder.
6. The user will be able either to save the logical set or also to save it and filter it.
7. As a result of the filter action a process will be run in the background creating a new itemized set with the relevant results of the logical set filtered with the conditions of the indication rule.
8. User will get an indication in the Recent Entities that a new itemized set was created or go to the set management to see the new itemized members.

Note: all users will permission to perform these actions since they are read only without the ability to change data

I've submitted the following to the Alma NERS 2020 process, which I hope captures both the spirit and specifics of this Idea, and welcome the support of all Alma users to also vote there.
Unlike Idea Exchange, with this Planned status from 1.9.17, NERS winners must be developed within a year according to IGeLU/ELUNA Product Development Agreement with Ex Libris.

Title: Improve Repository Search functionality with free text options for MARC tags and subfields in Advanced Search
Request ID: 6795
Description: Alma search function focuses on out-of-the-box mapping of grouped fields. It is a powerful search environment, but it does not allow for autonomy of any library staff to quickly and easily run a customised search by specific MARC tags and subfields on demand as need arises. For example, the Title grouping contains more than 3 dozen MARC tags and over a hundred subfields.
The current in-system solution is restricted to staff with a cataloger role, and is not quick or easy with several steps including a steep learning curve for writing complex indication rules. It also results in only static itemized sets.
This submission requests the addition of free text fields in Alma's Advanced Search, where any MARC tags and subfields for both bibliographic and holdings can be typed in, such as an All Titles set just on data in 245 a, 130 a and 730 a, or just in 260 b and 264 b, or just data in 776 z, or specific entries like LDR 06 equal to a and LDR 07 equal to m and 008 equal to 21 of d, or any combination together.
This functionality should be incorporated into existing Alma Repository Advanced Search options for all search types, including, but not limited to, All Titles, Physical Titles and Electronic Titles. It should have comprehensive conditions including “contains”, "not contains", “equals”, “not equals”, "is empty", "is not empty", and "starts with", as well as Boolean operator selection (AND & OR) between multiples lines of query, and the ability to select specific grouping of multiple lines with visual indication of this. The resulting records should be able to be saved as any Alma search set can currently, in a logical set, and also itemized immediately or later.
See mockup image provided and also Alma Idea Exchange entry with 264 votes as at 27.1.2020 and Planned status from 1.9.2017: Ability to search via free text MARC tag and subfields https://ideas.exlibrisgroup.com/forums/308173-alma/suggestions/17197337-ability-to-search-via-free-text-marc-tag-and-subfi

Would also love to know when the ability to search by MARC tags will be available.

Danette, that right there may be the most frustrating part of this whole thing. This is functionality we had in Voyager. Not in an ILS from another vendor. Not in some software that does granular searching/reporting and nothing else. In another ILS from Ex Libris. I realize the architecture is different, but clearly this is something they know we use, knew how to do in Voyager, and chose not to include in this "new and approved" software.

I agree that Yoel's solution does not fulfill what has been asked for in this idea submission. After reading the suggestion in Stacey's link, I'm not sure that this is the same thing, either, because it says it's asking for something other than creating sets. Maybe I'm reading it incorrectly?

We are part of a consortium, and not everyone has the same roles. That means that anytime a technician wants to create a set that searches data in a free text field, or any field not currently available in an advanced search index, we have to have someone with a higher-level role do the job for us. This is incredibly inefficient. These sets are not always needed to run a job, either - sometimes it is just for data collection. The "Search Builder" function in Voyager was fantastic for these types of granular searches, and the use of it didn't require any kind of administrative role. There are *many* times that techs at my library are stymied in our work by the inability to search the fields we need to search in order to create a specific set of bib records, holdings, or item records.

Does anyone know if the original idea submission request, which we all voted on, is going to be upheld as Planned?
The proposed plan in Yoel's comment back in September 2017 does not meet the original submission at all in focusing only on Holdings and still needing to use Indication and Normalization rules, rather than the clear request for free text Advanced Search functionality for both bibliographic and holdings records.
I note that a new idea was submitted with the same concern, which as of now has 74 votes: https://ideas.exlibrisgroup.com/forums/308173-alma/suggestions/14824554-ability-to-report-all-marc-tags-and-subfields-in-b
It seems clear to me what the community is asking for, and it is very disappointing if this feedback channel is being subverted by an erroneous Planned status.
Cheers, Stacey van Groll, University of Queensland

I agree with Mary. Indication rules are not a substitute for true indexing and searchabiliy/reportability. DG

In addition to searching the variable data fields, please include the ability to search the bibliographic/holdings LDR and 008 fields. There is a lot of valuable information coded in these fields. For example, in our holdings records, we routinely code receipt of acquisition status (6), method of acquisition (7), general retention policy (12), policy type (13), completeness (16), number of copies reported (17-19), lendy policy (20), reproduction policy (21), language (22-24), separate or composite copy report (25), and date of report (26-31). The bibliographic 006 and 007 fields should be searchable at the byte level as well.

Yoel's "solution" does not sound like it has much to do with this Idea. What we need is to be able to search any and all MARC fields and update any and all MARC fields, preferably without having to use indication rules since those are extremely complicated and the ability to creat and use them is not available to everyone at an institution. Yoel's posted "solution" does not do anything like this.

Does this mean that keyword indexing of MARC fields will be possible? I voted for this because I need to be able to search by MARC field and sub field to create sets of bib records - not holdings (although that would also be nice, obviously!)

Hello. We have read with great interest both the original description to this idea as well as the subsequent comments.
The original idea exchange description here is to have a search option "that allows us to enter the MARC tag number (and subfields) for searching, in both bibliographic and holdings records.".
At the current time we have decided to plan the option to create a set of holdings. In this way the staff user will be able to
1. create a set of holdings in the Alma user interface
2. filter the set with an indication rule (which filters by MARC fields)
3. if desired run a normalization rule on the filtered set to batch-edit it.
Further details regarding the time scope of the plan will be forthcoming

My wish relates to the holdings records :_ presently, the only possibility to search for an information is the tag 852 (subfields Call number and Classification). No search by subfields in tags 866, and notes tags exist. Could you please develop these possiblities?

Apparently indication rules do NOT work on holdings records. And there is MARC data in holdings records that we really need to search. Asking for each piece of data we need to search on a case by case basis is kind of nuts, and surely makes lots of extra work for ExLibris as well. It's a database. All the data should be exposed to searching and manipulation at the level of granularity it is entered. And for libraries right now that means MARC.

I wish we'd had this ability when we migrated. Would have made post-migration clean-up more efficient. Lots of stuff went into an "unknown" item type. Would have loved to be able to search on combinations of Leader, 007, 300 $b, etc. fields to figure out what was what. Would have saved alot of time going to the shelves to see if something in the AV area was a DVD, an audio cassette, a flimstrip, etc.

I just had a circumstance in which this would have been so useful! I needed to pinpoint a subset of dissertation records, and I eventually found a way to do so, but it would have been so much quicker and easier to do if this functionality were present in Alma. I know field/subfield combinations present in the records that I needed to find, but there was no ability to search them.

Thank you so much for considering this again. Having the ability to do true free text searching by MARC filed/subfield/indicator/position would make database quality control SO much easier and more efficient. Often those minutia buried in the fixed fields or other "not usually indexed" areas of the MARC record are the best, if not only, way to identify a specific subset of materials.

I fully agree with Lynne Billington's comment. Please do not forget to make different positions within MARC 008 searchable. As a serials cataloger I need to distinquish between different types of continuing resources both in search and analytics.