Talk:Photinia pageae

Did you know nomination

(

• Comment or view
• Article history

)

Photinia pageae leaf

• ... that 3 leaves of Photinia pageae (pictured) have been chosen for a machine learning image database?

• Source: Wilf et al 2021 "The dataset facilitates new research and education opportunities in paleobotany, comparative leaf architecture, systematics, and machine learning." & Appendix 1 "Rosaceae" entry for Photinia pageae

• Reviewed: Template:Did you know nominations/Diaphoretickes

Moved to mainspace by Kevmin (talk). Number of QPQs required: 1. Nominator has 375 past nominations.

Kevmin § 13:59, 29 July 2025 (UTC).

• Reviewing by list, as I know some people like this. Josh Milburn (talk) 15:59, 2 August 2025 (UTC)
  • ... that each boldlinked article is new enough. FINE
  • ... that each boldlinked article is long enough. The DYKcheck tool is helpful in evaluating these first two. FINE
  • ... that each boldlinked article is well-sourced, neutral, BLP-compliant, and copyvio-free. FINE
  • ... that each boldlinked article is presentable. FINE
  • ... that the hook is cited to a reliable source. FINE
  • ... that the hook is short enough. FINE
  • ... that the hook is interesting. This is the issue: I just feel at the moment I'm lacking context. I don't really like 'have been chosen' (it's passive) and it's not clear what being chosen 'for a machine learning database' means. I think the fact is a decent one; I just think it could be better framed! (Also, "three" rather than "3", I think.)
  • ... that any images are freely licensed, clear at a diminished size, and used in the article. FINE
  • ... that each QPQ has been done, where necessary. FINE
  • ... that there are no other, more subjective issues. FINE

@J Milburn: I worded with "have been chosen" due to the paper authors being very specific in the Fossils/fossil sites they took images from. The database itself is for AI lead and human lead pattern recognition to identify and classify fossil and living plants within known family frameworks.--Kevmin § 16:39, 2 August 2025 (UTC)

Sorry; I may have been unclear. My concern was with your use of passive voice, and the fact that the hook doesn't offer much context. (I.e., I have some amateur-level knowledge of both botany and machine learning, but I couldn't make much of it.) I was suggesting you try to rephrase it without the passive voice and with a bit more context. How about something like the following? Josh Milburn (talk) 18:50, 2 August 2025 (UTC)

... that [someone] included images of three leaves of the extinct plant Photinia pageae (example pictured) in an image database for [purpose]?

... that [database], a [description], includes images of three leaves from the extinct plant Photinia pageae (example pictured)?

@J Milburn: We run into the problem that DYK for several years has penalized or told nominators that unless a person can be blue-linked they are not to be included in a nomination (additionally the peer reviewed article makes it clear the database is a group collaboration of 17 coauthors. The database also was not given a name and combines multiple collections in a number of institutions. (Also we run into the 200 character limit by being to verbose fi we have to lay out that amount of detail, while hooks are recommended to be short and catchy).--Kevmin § 22:52, 2 August 2025 (UTC)

They were just recommendations; if the database doesn't have a name or we can't mention the scientists by name, then perhaps we can approach it another way. I agree that a hook should be short and catchy, but they also need to be understandable. Per WP:DYKINT: 'Make sure to provide any necessary context for your hook; don't assume everyone worldwide is familiar with your subject.' Josh Milburn (talk) 07:20, 3 August 2025 (UTC)

@J Milburn: That is why I chose the wording I did. if you have been following the DYK:talk page you will have seen the volume of controversy constantly flowing around invocation of DYK:INT (and where I stand on something that is hypersubjective when the mainpage receives over 25 million views in a 12 hour run period). If I wordsmith I would tweak the hook as such:

Alt1 that 3 fossils of the rose relative Photinia pageae (pictured) have been chosen for inclusion in an AI and human learning database?

this is a little more wordy but not deviating from the actual source material at all.--Kevmin § 14:04, 3 August 2025 (UTC)

No, I haven't been following conversations on that particular talk page. If you feel it's important I know something, you're welcome to tell me. Clearly, the guidelines at WP:DYKINT require judgement. But that does not make them 'hypersubjective', and nor does it make concerns that appeal to them unactionable. I appreciate that you've made an effort to come up with a new hook, but, as far as I can see, you've overlooked two of the concerns I raised, which were nothing to do with the 'subjective' guideline, but everything to do with plain writing. Could I suggest the following, which I've based on ALT1? I dropped 'human learning' because it was getting a little wordy. Josh Milburn (talk) 15:11, 3 August 2025 (UTC)

ALT2: ...that the creators of a botanic database for machine learning included three images of leaves from Photinia pageae (example pictured), an extinct relative of the rose?

I am okay with Alt 2.--Kevmin § 15:58, 4 August 2025 (UTC)

for Alt2. Josh Milburn (talk) 18:45, 4 August 2025 (UTC)