As I mentioned I don't really have a need for this manual rating system, although I think I can see some benefit. So I'm really not asking for this feature, but since you (Taka
) thought that this was "…an interesting idea" and made the thread sticky, it sounded like it was something you were considering. Then when zakky
requested the ability to delete items and it sounded like you weren't considering implementing his request, my brain connected to the two requests as two use cases of a manual item-based rating system (combined with the ability to filter items above or below a certain threshold and the ability to identify channels above or below a certain threshold
Since it seemed like you were somewhat interested in this I thought I expand upon the original request and demonstrate how you might be able to satisfy zakky
's request at the same time.
The fact that no one else has replied to KevinP
's post for 10 months indicates that his request must be "niche need". If 500 paying Awasu customers had replied and said they wanted a manual rating system, and 500 free Awasu users said they would upgrade if there were a manual rating system, I pretty sure you would be convinced that they're a need for this feature.
"…you haven't covered… …what would such a feature be used *for*?"
Actually, I think that I (and the original posters
) did outline two uses:
's desire to find channels that he has read, but didn't find high quality content. He may want to delete these channels.
's desire to "delete" items from search results. This would have to be a new category of Channel Filter criteria.
Outside of these two use cases I'm not sure what it could be used for.
"People don't use Awasu for the sake of using Awasu."
Neither the original posters (OP) nor I meant to imply this. Both OPs mentioned a need, and I was just trying to connect their requests. However people do use Awasu, and in KevinP
's case it sounds like he expressed an Awasu Channel management need.
"…my guess would be that a large percentage of the articles will be ignored and only a few rated, thus giving skewed results"
I 100% agree with you that most items will not be rated and thus will have the default rating only, whether that skews the result or not, I'm not as sure. With a 1-5 rating scale and a default value of 3 (neither good nor bad), if KevinP
rates one item as a 2 in a channel of 10 items, using a simple Average function the channel's rating will be 2.9; in a channel of 100 items with a single item rated 2 the channel's rating will be 2.99. If he can run a "Channel Management" report to find all channels whose average item rating is below 3 he'll see both channels (hopefully with their score displayed) on the report even thought he's only rated 2 items. He'll see more dramatic results with the more items he rates and if he rates them with a 1 instead of a 2.
"…how is it going to be remotely manageable?"
As demonstrated above a user wouldn't have to rate everything as long as there's a sensible default value for all unread items and the numeric precision of Awasu is finite enough to discern 2.99999 from 3.
"…minimize how much stuff you have to do manually"
Both of the OPs requested a manual feature, and you just added a manual "Send to" feature in version 2.4, so I don't think their requests are totally out of line. By the way a big thank you for all the new Send to options particularly the ability to define my own.
To a certain degree, manually sending an item to a Workpad is a positive vote of confidence (a type of manual rating
) for an item, but there isn't a way to track it over time and see: "Which Channels did I find the most Workpad-worthy items?" and "Which Channels have I never found a Workpad-worthy item?"
In fact KevinP
's request (and maybe zakky's request
) could almost be satisfied with Send to tools, but that's another post for another day (stay tuned
NewsGator's Attention Report
They way I manually use Awasu is as you mentioned ("…skim through large amounts of information quickly…"
) and I "Mark all items as read", doesn't translate well to NewsGator's "Attention Report". If I open a channel and skim it quickly, I've given it my "attention" and NewsGator's algorithm will increment the channel's attention counter, even though I spent less than 10 seconds in the channel. That doesn’t work for me; it provides many "false positives".
Yes I'm not overly trilled by most of the "if you liked 'that', this then you'll love 'this'" recommender systems. Although I have found them useful in some cases, so I think the important thing is that they do "get out of your way" when you don't want to use them, but are on demand when you do.
I don't get it either. Taka, I'll read you blog, but if you start "tweeting" don't expect me to subscribe.
I won't beat this dead horse anymore; it wasn't even my horse to begin with. I'll make my own feature requests at some point in the future, probably more generic API kind of things. Thanks again for the work you've put into Awasu to enable us to "hook into the flow of information" and perform our more-specific tasks easily.