Open wsanchez opened 5 years ago
From @flwyd:
I think I understand what you're saying. I might characterize this as "Recommended related incidents," calling to mind things like YouTube's recommended videos or Amazon's related products features. The term "pop up" to me sounds like a dialog box that gets in your way, which sounded distracting. But if there was a section at the bottom or off to the side of the incident view that showed a few potentially related incidents, that might create an opportunity to create historical situational awareness. This could be based on structured information (e.g. location, incident type) or soem kind of language processing that identifies uncommon words or phrases (though we'd want to make sure that it doesn't recommend all of Hubcap's other incidents just because Hubcap called this one in…).
Tool (@wsanchez): thoughts?
Tool (@wsanchez): thoughts?
Sounds nice. :-)
Probably more work than is likely to happen, though. Right now IMS doesn’t have a notion of “related incidents”, though that is an open ticket which I do intend to look at (for duplicate incidents, at a minimum).
If we had a massive system with tons of data, that sort of thing is a lot more worthwhile. Yes, we depend on people knowing things right now, but the IMS software having that kind of smarts is a big level up.
Perhaps my Machine Learning game needs a boost, but right now, we don’t have that expertise available.
Actually, re-reading @flwyd: reply, I think that he's right about the UI, and if we can find some code someone that can do some "this thing looks related to that thing" magic, perhaps this is doable.
Filed for Apparatus:
IMS pop-up feature when certain keywords start to reoccur in the system a certain number of times
What am I looking for?
What I have in mind would be some kind of feature in the IMS where if the system sees us entering words or phrases (probably mostly nouns) on a repeating basis a small pop-up window opens in like the bottom corner of the screen to suggest possibly linked incidents. So If I start keying in the phrase"White Ocean" the pop-up would say "Possibly related incidents" and would have links to the other IMS entries with White Ocean mentioned. Recurring camp names, locations, mutant vehicles, participant and other staffer names, installations, etc, all of these be helpful to have suggested links pushed to a small window for various other reasons. Even if certain locations like the Stick always get suggested, those could just be ignored or not based on human discretion. I imagine that the Operators would mostly be the Rangers who would see these pop-ups and peek at the potentially related incidents and pipe up to Khaki with, "Yeah, that's the fourth time we have been out to that camp for similar incidents." or somesuch and then we would know that there were related incidents and we could use that information as related context for the current incident.
Why is this useful? To help us make connections that we might not already know to look for. Right now as a Shift Command person I would have to already know or have some reason to suspect that White Ocean had prior problems before I would think to search for it. Right now we rely on the hope that we will always know what important things will come up again and that they got properly included in the shift change pass-downs. Even the important stuff drops off our radar after a day or two usually.