Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Bought Matcha for Its Content Recommendation Algorithms
- Thread starter MacRumors
- Start date
- Sort by reaction score
First they had iTunes Match.
Now they could have iTunes Matcha. It's that bit matcha than match.
not a good addition, i dont know any people who use recommendation apps, they are useless, channels are doing enough marketing about their serials, but i can say that more goodies never hurt.
It's absolutely essential that we get a better way of finding shows to watch.
Sure we can let a channel develop a reputation for certain types of shows, and let it present the next show when we finish watching one, and hope it fits our interests ... and if we all like the same kind of things that would work. But having the stb help us find a small number of shows we would genuinely consider will be very powerful.
If you think about it the App Store is brilliant for Apple. Just sit back and see what the best App is, see what people think about it, then buy the company!
absurd and unnecessary. just beam out via AirPlay to the Apple TV.
----------
how long did it take Siri to go from app to iphone4? little over a year. oh nos! I can feel my life force...draining. apple owes you so much...
----------
how do you figure that? Siri took 15 months to go from app to OS. do you have any real data like that to base an opinion from?
----------
that isn't how software works. without access to their source code, apple (or anyone) can only attempt to create similar functionality.
nobody thinks code shouldn't be copyrightable. copyright is very different than patent. copyrights protect your actual code/novel/whatever from direct copied plagiarism. patents protect vague ideas.
----------
If I were one of the Siri guys I'd be pretty happy to see my idea in millions and millions of devices.
also -- the point of starting most IT companies is to attract investors and sell. it's about making money in an enjoyable way.
not slow enough to stop you from seeing the headlining, clicking the link, reading the story, and posting about it. mission accomplished.
Precisely the kind of laser-focused acquisitions, Apple claims to be aiming for.
In unrelated news, after their work places (Utrecht Netherlands, Red Light district) had been closed by the government because of suspected human trafficking by pimps, a group of prostitutes has started a collective called 'Macha' (for 'female Macho') and requested a permit to start working again. The collective plans to hire its own protection and share all revenue and cost.
agreed...but the recommendation apps arent good enough, they link grey's anatomy to two and a half men, they link dexter to desperate housewives etc.
there's in no TRUE recommendation app right now.
Yes they use general associations, it seems. There can be significant viewer overlaps between two very different shows, but that's not enough - the systems need to realise that they are quite different, and take into account that what I personally like about one show might not be in the other show.
Our brains are wired for rewards and punishments with 2 separate systems, why wouldn't that go into film and TV preferences? I don't like films where characters react in ways that don't suit their character - I want a realistic reaction. A film can have everything I like in it, but if it also has something I don't like it'll put me off.
I'm also a fan of correlating films someone does NOT like, instead of what they like. "Greg doesn't like Two and a Half Men, he probably also won't like "2 broke girls".
But the better method is to have a system that looks at what you like, and finds common threads of what makes you think a show is good, and also finds threads of what you find bad, so it can predict how YOU would score a show, not just how most people who like Greys Anatomy would score a show. I still think Jinni does that pretty well, they just don't have a search system that actually highlights the films you would rate as "awesome", seems an odd omission!
It's all complicated, but extremely worthwhile, and I hope Apple is getting there somehow
Yes they use general associations, it seems. There can be significant viewer overlaps between two very different shows, but that's not enough - the systems need to realise that they are quite different, and take into account that what I personally like about one show might not be in the other show.
Our brains are wired for rewards and punishments with 2 separate systems, why wouldn't that go into film and TV preferences? I don't like films where characters react in ways that don't suit their character - I want a realistic reaction. A film can have everything I like in it, but if it also has something I don't like it'll put me off.
I'm also a fan of correlating films someone does NOT like, instead of what they like. "Greg doesn't like Two and a Half Men, he probably also won't like "2 broke girls".
But the better method is to have a system that looks at what you like, and finds common threads of what makes you think a show is good, and also finds threads of what you find bad, so it can predict how YOU would score a show, not just how most people who like Greys Anatomy would score a show. I still think Jinni does that pretty well, they just don't have a search system that actually highlights the films you would rate as "awesome", seems an odd omission!
It's all complicated, but extremely worthwhile, and I hope Apple is getting there somehow
yep, i also hope that apple brings out a TRUE recommendation app
Haha touche.
----------
I read the article in hopes of gaining substantial new info. Sadly that didn't happen, so I posted a snarky comment in my frustration.
This is a great site, gives good info, aggregates stories from all over the web. We hold it to a standard. Only a few front page headlines are written daily, so each story holds a good amount of weight with its readers.
When an article is basically a regurgitation of something that's already been covered, and very recently, it's annoying, and diminishes the value of the site.
A simple update to the old article stating that the reasons rumored for the acquisition were confirmed would have sufficed and been more in line with how the site normally operates.
I just take as many articles as I can. Like you said, there are only a few a day. I need MOAR.