[Discovery_PVE_Comm] Secret search parameter to improve title search results?

Tue May 21 14:44:57 UTC 2024

Hello,

I'm very much in favor of contacting Ex Libris to enable this feature in our sandboxes. Like Lindsey mentioned, I would be keen to test this with boosting 245 $a.

If we have it enabled by our next meeting, we can make testing it a priority for our July meeting.

All best,

Andrew Belongea
Metadata and Systems Librarian
Columbia College Chicago Library
abelongea at colum.edu<mailto:abelongea at colum.edu>
(312) 369-7884

________________________________
From: Discovery_PVE_Comm <discovery_pve_comm-bounces at carli.illinois.edu> on behalf of Skaggs, Lindsey via Discovery_PVE_Comm <discovery_pve_comm at carli.illinois.edu>
Sent: Monday, May 20, 2024 12:26 PM
To: Laura Spradlin <lspradli at iwu.edu>; Gibson, Jessica <oram at uillinois.edu>
Cc: Discovery Primo VE Committee <discovery_pve_comm at carli.illinois.edu>
Subject: Re: [Discovery_PVE_Comm] Secret search parameter to improve title search results?

Caution: This is an external email and may be malicious. Please take care when clicking links or opening attachments.

I’m intrigued – I think this could solve some of my specific search frustrations and I’m on board with enabling it in the sandbox for further consideration.

In reading the thread, it looks like we’d boost 245$a specifically. I’ve mentioned in meetings my frustration with “Harvard Business Review.” Other titles regularly outrank the journal title despite my attempts at boosting (leading to implementing recommended resources):

[cid:image001.png at 01DAAAB0.2AD63BA0]

For example, in the first result below, “Selling in foreign markets,” the phrase “Harvard Business Review” shows up in the 130$a, 245$c, $490a, and $500a which outranks the 245$a and 222$a entries in the journal record. My read is that this setting change would ameliorate that.

Thanks,

Lindsey

From: Discovery_PVE_Comm <discovery_pve_comm-bounces at carli.illinois.edu> On Behalf Of Laura Spradlin via Discovery_PVE_Comm
Sent: Friday, May 17, 2024 3:41 PM
To: Gibson, Jessica <oram at uillinois.edu>
Cc: Discovery Primo VE Committee <discovery_pve_comm at carli.illinois.edu>
Subject: Re: [Discovery_PVE_Comm] Secret search parameter to improve title search results?

This message originated from outside of the Illinois State University email system. Learn why this is important<https://urldefense.com/v3/__https://help.illinoisstate.edu/technology/page-88179827.html__;!!DZ3fjg!9eYg_w78sg3PqKC8YtprQeXeXpqhODKCVHTRhum_m8P5L_Fi7kQ30Zug6lOwWFJxmrhMSnJ7w_6m7ea4PR6J44z-U-w3jyrZEUU$>

Thanks for sending this, Jessica! Happy to have wrenches thrown in, especially if it can improve things broadly!

Sorry for the long line of screenshots below, but I think this might be a relevant example--one use case I've been playing with is The Beet Queen by Louise Erdrich. Searching beet queen prompts Primo VE's search expansion with controlled vocabulary and includes "Louise Erdrich" in the search. The Beet Queen comes up as result #5 in an All I-Share Libraries search (boosting did slightly improve this; previously, our IWU holding, which is grouped in the #5 result, was #7).

[cid:image002.png at 01DAAAB0.2AD63BA0]

However, if I click "Just search beet queen," The Beet Queen comes up as the #1 and #2 result (boosting also helped our copy rise to the top--our IWU holding is in the first 6-version FRBR group, which, prior to boosting, was the #7 result; the second 3-version FRBR group, which does not include our holding, was #1):

[cid:image003.png at 01DAAAB0.2AD63BA0]

If I search the beet queen, the search expansion automatically comes up, but now The Beet Queen comes up as results #1 and #4. Our IWU holding is in result #4:

[cid:image004.png at 01DAAAB0.2AD63BA0]

And if I click "Just search the beet queen," the 3-version FRBR group is the #1 result with the 6-version FRBR group at #2:

[cid:image005.png at 01DAAAB0.2AD63BA0]

So, in this example, it seems like "the" is having an impact, but so is the search expansion (which we have not done any real testing on).

Based on the descriptions in the listserv messages, it does seem like this setting would help and be worth testing, if for nothing else than it doesn't seem like "the" should make that much of a difference in search results.

Thank you!
Laura

Laura Spradlin (she/her/hers)
Electronic Resources & Systems Librarian

Assistant Professor, The Ames Library<https://urldefense.com/v3/__https://www.iwu.edu/library/__;!!DZ3fjg!_elE7my7R2jIEv9qwQBHndMQ9iPLgpXEoBRZAtcLvfG--KuSrkeXBpWi4KztVuIKFyklxevY85QYiTYki8Od2HEVmc1on7S0s0_M$>
Illinois Wesleyan University

(309) 556-3144

Office Hours & Liaison Areas<https://urldefense.com/v3/__https://libguides.iwu.edu/prf.php?account_id=329085__;!!DZ3fjg!_elE7my7R2jIEv9qwQBHndMQ9iPLgpXEoBRZAtcLvfG--KuSrkeXBpWi4KztVuIKFyklxevY85QYiTYki8Od2HEVmc1onzhh9E0M$>

On Fri, May 17, 2024 at 10:18 AM Gibson, Jessica via Discovery_PVE_Comm <discovery_pve_comm at carli.illinois.edu<mailto:discovery_pve_comm at carli.illinois.edu>> wrote:

Hi, everyone,

While I don’t want to throw a wrench into Laura’s upcoming part of the June 11th webinar on Search Ranking, I’ve been catching up on my Alma/Primo customer list emails, and I found the recent thread (attached; read from the bottom up; the magenta highlighting is mine) to be quite interesting. Apparently, there is a previously secret customer parameter, discovery_customer_defined_exact_match_field, that might help with exact title search matching.

I would like to get the Committee’s thoughts on this since it sounds like it may be another on setting that needs to be coordinated consortially, but our next meeting isn’t until June 5, less than a week before the webinar. Could we at least start to discuss via this topic via this committee email list in advance of our meeting?

Are you all experiencing the same issues outlined in this thread with searching?

Do you think this setting might help based on the descriptions?

Should I file a Case to ask Ex Libris to see if they will enable this setting in our Sandbox so we could compare the differences in our 5 sandbox IZs and NZ Primo VEs with our production Primo VEs?

We can perhaps also do some basic testing using other sites Primo VEs who indicate they have turned it on; for example, Lesli’s library is https://urldefense.com/v3/__https://thslc-houston.primo.exlibrisgroup.com/discovery/search?vid=01TEXASHEALTH_HOUSTON:TMC_INST&lang=en__;!!DZ3fjg!6J3_XXQrl8K11bJQoM6tDShT8aIINF7sGZSwxTozXTl_0TY-MdIAWBrdjwLX0TNhID7324KfrKwPajcJqyG7kKeW_W_d3LDsDfcg6fI$ <https://urldefense.com/v3/__https://thslc-houston.primo.exlibrisgroup.com/discovery/search?vid=01TEXASHEALTH_HOUSTON:TMC_INST&lang=en__;!!DZ3fjg!_1S9731apWaL8OW9Xm5SFXDd-5nleteqvWltEcEskSgSikhSuyYMu7_33hm8LegHRWQHzCMcXP1SKXmSvLHh46PknHLLitBSNpY$> .

thanks,

-Jessica

--

Jessica Gibson

Assistant Director, System Services

Consortium of Academic and Research Libraries in Illinois (CARLI)

--
Discovery_PVE_Comm mailing list
Discovery_PVE_Comm at carli.illinois.edu<mailto:Discovery_PVE_Comm at carli.illinois.edu>
https://lists.carli.illinois.edu/cgi-bin/mailman/listinfo/discovery_pve_comm
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.carli.illinois.edu/pipermail/discovery_pve_comm/attachments/20240521/eb9028e7/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.png
Type: image/png
Size: 36468 bytes
Desc: image001.png
URL: <http://lists.carli.illinois.edu/pipermail/discovery_pve_comm/attachments/20240521/eb9028e7/attachment-0005.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.png
Type: image/png
Size: 129753 bytes
Desc: image002.png
URL: <http://lists.carli.illinois.edu/pipermail/discovery_pve_comm/attachments/20240521/eb9028e7/attachment-0006.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image003.png
Type: image/png
Size: 105490 bytes
Desc: image003.png
URL: <http://lists.carli.illinois.edu/pipermail/discovery_pve_comm/attachments/20240521/eb9028e7/attachment-0007.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image004.png
Type: image/png
Size: 112173 bytes
Desc: image004.png
URL: <http://lists.carli.illinois.edu/pipermail/discovery_pve_comm/attachments/20240521/eb9028e7/attachment-0008.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image005.png
Type: image/png
Size: 106589 bytes
Desc: image005.png
URL: <http://lists.carli.illinois.edu/pipermail/discovery_pve_comm/attachments/20240521/eb9028e7/attachment-0009.png>