Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Performance problem in "recommend by author" plugin #5887

Open
asmecher opened this issue May 15, 2020 · 10 comments
Open

Performance problem in "recommend by author" plugin #5887

asmecher opened this issue May 15, 2020 · 10 comments
Assignees

Comments

@asmecher
Copy link
Member

asmecher commented May 15, 2020

The "Recommend By Author" plugin queries for an author across all journals in the installation, in all supported locales. With large numbers of authors and more than a few locales, this can get quite slow.

The issue was reported (by @ajnyga) in OJS 3.1.2; in OJS 3.2.0, the query appears to have been rewritten to be journal-constrained, which is probably not intentional.

  • Fix performance issues with large data sets
  • In 3.2.0, verify that the plugin correctly lists recommendations in other journals on the same installation
@asmecher asmecher added this to the OJS/OMP/OPS 3.2.0-4 milestone May 15, 2020
@asmecher asmecher self-assigned this May 15, 2020
@asmecher
Copy link
Member Author

@asmecher asmecher removed this from the OJS/OMP/OPS 3.2.0-4 milestone May 20, 2020
@mfelczak
Copy link
Member

For a multi-tenant install or an install with unrelated journals, it would be preferable to keep this plugin to be constrained by journal or to add a setting to retain this functionality.

@ajnyga
Copy link
Collaborator

ajnyga commented Feb 19, 2021

Was this fixed for 3.3?

@asmecher
Copy link
Member Author

No, I don't think there's been any dev activity on this one yet.

@asmecher
Copy link
Member Author

@mfelczak, this feature currently searches OJS's author records by given and last name, looking for exact matches. This performs badly but also doesn't align with our plans for author disambiguation using ORCIDs. Do the clients who are asking for this feature tend to use ORCIDs?

@mfelczak
Copy link
Member

mfelczak commented Sep 7, 2021

Hi @asmecher, I don't have a firm sense of ORCID adoption across our hosted journals as it's not something that we currently track. We could provide some stats though if that might be helpful, e.g. checking whether the ORCID plugin is enabled.

@asmecher
Copy link
Member Author

asmecher commented Sep 7, 2021

Rather than stats for existing users, I'd be curious whether editors think adopting it in place of "recommend by author" would meet that feature's need. At best, an OJS-based author list can only show the results in that single installation; using ORCiD, it'll be a global list of everything that author has written. I suspect end users interested in following an author would much prefer the latter. Editors might want the former as a "brag page" but I'm not sure it'll help researchers.

@NateWr NateWr moved this to Backlog in Infrastructure May 9, 2022
@jonasraoni jonasraoni self-assigned this Mar 1, 2023
@jonasraoni
Copy link
Contributor

I can confirm this plugin is a bit resource intensive. As I've a good place to test some improvements, I've assigned it to me.

@Ian-BD
Copy link

Ian-BD commented Oct 23, 2023

Has there much headway in improving the performance of this plugin?

We run a large multi-journal site and was having some major performance issues which we've narrowed down to this.

For now we've disabled the plugin but if you have any suggestions as to how we can make some temporary workarounds while you work on improvements that'll be greatly appreciated.

@jonasraoni
Copy link
Contributor

@Ian-BD I've just had time to update a similar plugin (#8710). Last time I've skimmed the code, I saw some opportunities to optimize it, in case you get it fixed, feel free to contribute :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
Status: Backlog
Development

No branches or pull requests

5 participants