Stacey van Groll
My feedback
148 results found
-
7 votes
An error occurred while saving the comment -
92 votesStacey van Groll
supported this idea
·
-
86 votes
An error occurred while saving the comment Stacey van Groll
commented
I see the idea has been updated to delete the Closed status. If Ex Libris is watching flow-on correspondence here, it would be appreciated if there would be more transparency in manner of communication. This is easily done by leaving the mistake visible and adding another status over it, which has been done in other ideas such as in the Content forum. This is vastly preferable than hiding something and pretending it never happened, with the resulting impression of not being open and trustworthy.
Stacey van Groll
supported this idea
·
-
3 votes
An error occurred while saving the comment Stacey van Groll
commented
There is Code of fulldisplay.Access_content_in with Description option such as:
View online: {{provider}}This would seem to lead to the desired outcome if changing the Description accordingly in your environment per the example screenshot:
Click here for access: ProQuest Central -
1 vote
An error occurred while saving the comment Stacey van Groll
commented
No offense intended if not, but is this a bot?
It reads strangely and some of these suggestions would not be suitable and disruptive to case management for how cases might be handled by each site (Automatic Case Creation) or reference aspects of SalesForce customers can't see (Combined Dashboard). -
3 votes
An error occurred while saving the comment Stacey van Groll
commented
I’ve recently been troubleshooting an issue with search results on a provider platform which reminded me of this submission.
I already had concerns with it in anticipating it causing search return problems in Primo, but this analysis I conducted on the other platform solidified the issue in my mind, and I thought it useful to share here as to why I would not support this submission.The article in question:
How to make sure you’re not using data just to justify decisions you’ve already madeThis issue was that results for an exact title match were not returning the known item, unless entered with quotation marks.
I noticed that in the results which were returned, the term highlighting for matches to the search query stopped after “you’re” even though “not” also matches the query.
Given “not” is a known Boolean operator, I suspected that there might be an issue where it’s being treated incorrectly to tell the search engine to not return results.
Hence Boolean operators in capitals typically to avoid this problem. And this is the case for Primo with NOT, AND, OR.
Adding quotation marks worked to return the known item though because they force an exact phrase match, as opposed to Boolean behaviour, and overrides the defect for “not” being treated as Boolean.
And sure enough with some iterative testing it is found that skipping the word “not” when not using quotation marks resulted in the desired article being returned.• Query: How to make sure you’re not using data just to justify decisions you’ve already made – Results: 22 results and the desired article is not among them, despite being an exact match
• Query: How to make sure – Results: 783 results and the desired article is not in the first 50, but presumably it is in the results set somewhere
• Query: How to make sure you’re – Results: 27 results and the desired article is No.6
• Query: How to make sure you’re not – Results: Same as directly above, with 27 results and the desired article is No.6
• Query: How to make sure you’re not using – Results: 22 results, and the desired article is missing (repeat for additional words added). Assumption here being that the desired article is not returned because the “not” tells the search engine that the results need to contain “How to make sure you’re” but NOT include “using”
• Query: “How to make sure” – Results: 388 results and the desired article is not in the first 50, but presumably it is in the results set somewhere
• Query: “How to make sure you’re” – Results: 23 results and the desired article is No.5
• Query: “How to make sure you’re not using” – Results: 5 results and they are all the desired article. Assumption here being that quotation marks are overriding Boolean to force an exact phrase query
• Query: How to make sure you’re using – Results: 5 results and they are all the desired article -
86 votes
An error occurred while saving the comment Stacey van Groll
commented
It should be standard to add fields like this to Analytics in alignment with search indexes in Alma. Ideally at the time of development, but certainly when they are revisited at a later time when customers flag an important field which is lacking.
There are other Notes fields ETL'd to Analytics and performance issues should be controlled well by the existing Analytics character limits. -
35 votes
An error occurred while saving the comment Stacey van Groll
commented
I'm surprised this would be considered an enhancement request. I can't think of other jobs where the job name is not clear to state the job which failed with granularity by profile. Was this advised in a case that it was an enhancement rather than a defect?
-
0 votes
An error occurred while saving the comment Stacey van Groll
commented
We have more than one publishing profile using OAI-PMH and just use different sets, so I believe this is possible already.
-
18 votes
An error occurred while saving the comment Stacey van Groll
commented
I'm not sure I understand completely, but the full text aspect is controllable by the site to activate the collection. For example, an Unpaywall collection is available as of a few months back to resolve a lot of these instances. I personally would be concerned to undermine the premise of the default search experience including full text available resources, if there is no full text link. Otherwise, if Open Access resources with a full text link are not included in the default search, that sounds like a defect. Perhaps provide an example to help resolve this confusion as to the meaning of this submission?
-
14 votesStacey van Groll
supported this idea
·
-
6 votes
An error occurred while saving the comment Stacey van Groll
commented
I've noticed student queries seeking ability to limit results by place. Use cases include higher degree students for a specific research focus, or undergraduates who have assignments where their lecturers stipulate that they need to include national or regional emphasis.
It's certainly an issue to have a facet like this which only has local results, as I can imagine patrons being upset with potentially being misled when using it and not realising that they are excluding all CDI results even though publication information is visible in the records.
It might still be difficult also though to make such a facet meaningful and useful, given the place of publication could be very granular to a particular state, when patrons might be interested in a country. That would be onerous to manage for selections, particularly with facet value limits.
Overall though, it would be great to have the option of a blended facet for this, for records from all data sources including CDI. -
2 votes
An error occurred while saving the comment Stacey van Groll
commented
I'm concerned that this would be considered an enhancement request when it's clearly marked as a known issue ie a defect. I suggest submission as a SalesForce case, to encourage Ex Libris to action a fix.
-
6 votes
An error occurred while saving the comment Stacey van Groll
commented
I'm not quite sure if this is meant to be two separate submission? For the second portion, I wonder if you've cleared cache/cookies recently or not allowed functional cookies? This may cause Recently selected entries to not appear as expected.
Otherwise, that sounds like a regression (which describes functionality which used to exist and now doesn't due to an issue), not an enhancement request. -
2 votes
Thank you for your suggestion. The Rialto team is reviewing this idea to determine how it might fit into our future plans. We cannot provide a timeline for these ideas, but be sure to check back often and vote for the ideas you support to receive status and comment updates.
Best,
Heidi Whitehead
Rialto Product Manager
Stacey van Groll
shared this idea
·
-
3 votes
An error occurred while saving the comment Stacey van Groll
commented
This would be a defect according to the documentation, which is clear that the setting works on a collection level.
Could it perhaps be the case that the articles are coming from another source which is not suppressed? You can check this in the CDI Activation Analysis Tool.
This setting will not prevent continuing presentation unless you suppressed all of the collections containing the article, as it is collection specific (and this is documented).
The permalink IDs may still include the suppressed collection, but that's because it's a randomised ID from the participant records. -
6 votes
An error occurred while saving the comment Stacey van Groll
commented
Have you tried the Partner setting for "Enable service for guest user - Indicates whether the link to the partner displays in Primo when the patron is not logged in."?
We don't have this active usually, as we offer resource sharing only for certain logged in patrons, but I tested it in Sandbox and it correctly caused the option to appear in Primo Get It. -
69 votes
The Rialto team has reviewed this idea and this work has been planned on our short-term roadmap. Thank you for your feedback and support.
Best,
Heidi Whitehead
Rialto Product Manager
Stacey van Groll
shared this idea
·
-
3 votesStacey van Groll
supported this idea
·
-
299 votes
An error occurred while saving the comment Stacey van Groll
commented
I also had a case for this opened in May 2020 where I directly referenced this Idea Exchange submission, and will copy details relevant comments from my case log.
In sum, I think this was a defect that was fixed in August 2022 and will reiterate my last comment that I believe anyone still seeing this should open a case.
CDI - Different term highlighting in record display does not match query, in comparison to PCI where highlighting is complete including stop words
* 05310266Analyst advised: "In case of an exact match, the title highlight should include also the stop words. This, as you mentioned, is a good way to let the user know easily they hit the right result. If the match isn't exact, the stop words shouldn't be highlighted, again in order to allow the user a quick initial navigation among the brief results. But there are are few variations to this - namely, what counts as "exact match". For example, the search for "Playing the game; is sport as good for race relations as we'd like to think" - will get the full highlight. But it can be flexible - if you change some of the words in the query to their plural/singular forms ('races', 'sports', or 'relation') you will still see the full highlight. However, CDI isn't expanding 'we' when the indexed term is 'we'd'. Development team verified that this is the expected behavior for CDI, and that in fact it was the same for PCI (which is why 'we' wasn't highlighted then). So, the basic behavior is the same." 28.4.2021 SvG
Responded to query only item outstanding re the Idea Exchange submission included in the original case for random stop word highlighting. 29.4.2021 SvG
Analyst advised: "Currently, there's a fix planned for both for CDI and Primo side, to prevent the highlight of stop words when they appear separately, in addition to the expected highlight of them as part of an exact match. For example, a search for "lord of the rings" will highlight, as expected, the full exact match, but also the words "off" or "the" before or after the full highlight.", and also seemed to not see the issue in the Idea Exchange entry. 29.4.2021 SvG
Responded to point analyst to screenshot on the entry, which shows the issue. 30.4.2021 SvG
Analyst confirmed screenshot missed, and that it seemed an issue with the separate stop words as per the planned fix and not the "exact match" highlight. 25.2.2021 SvG
Analyst advised tentatively planned fix for May 2022, which is extended as the fix must include also changes to the search mechanism. 15.6.2021 SvG
Analyst advised fix estimation pushed back to August 2022. 20.2.2022 SvG
Analyst advised firm fix in August 2022 Release. 8.7.2022 SvG
Update of fix in the August 2022 Release and case set to closing. 22.8.2022 SvG
Responded to advise cannot confirm and added a screenshot per the described fix in April 2021 showing the same additional highlighting. 22.8.2022 SvG
Analyst is querying with PM re: "stop words detached from the query itself are no longer highlighted, but the adjacent stop words are. " 22.8.2022 SvG
Added comment referencing earlier updates by analyst and the desire to ensure both aspects are resolved: "to prevent the highlight of stop words when they appear separately, in addition to the expected highlight of them as part of an exact match. For example, a search for "lord of the rings" will highlight, as expected, the full exact match, but also the words "off" or "the" before or after the full highlight." 24.8.2022 SvG
Analyst advised: "What was actually fixed in this case is that only the words in the search phrase will be highlighted, not only the 'exact match'. For example, when searching 'lord of the rings', these words will be highlighted for all phrases that contain these words adjacent to each other. Indeed, according to the release notes, previously, the text 'of the' was highlighted even though it did not begin with Lord and/or end in Rings and currently, 'of the' is highlighted only when it appears in adjecent to the search phrase." 8.9.2022 SvG
An error occurred while saving the comment Stacey van Groll
commented
I still cannot replicate this. I'll attach images of what I see for one of the last examples Lacey posted with custom code stripped out to clearly show no stop word underlining.
I suggest this be opened as a case for anyone seeing it, so there can be explanation provided for what appears to be an intermittent defect.An error occurred while saving the comment Stacey van Groll
commented
I'm not seeing 'the' highlighted in my Primo, Lacey.
Can you provide a link to your environment?
Is this asking for an enhancement to hide the Resource Sharing Service if the site has enabled Unpaywall PDF links, and the record shows both?
Generally my understanding is that this only happens when the site hasn't also enabled the Unpaywall collection for full text links, and then the API to show the Unpaywall PDF link is returning a result but the CDI collection is a little bit behind in being up to date.
The example isn't returning a result.
I am also wondering why this is referring to "scheduling of a fix" when defects should not be added to Idea Exchange, which is for enhancements.