You're watching...

Can Twitter predict who will win the Oscars?

Details

  • Description

    Steve Canepa looks to Twitter predict Oscar wins

  • Duration 5:52
  • Date

Clips

Also in this playlist...

Latest Entertainment

Auto-advance: ON

Auto-advance

Transcript

This transcript is automatically generated

Steve Canada and general manager of media and entertainment at I Deanna says Steve thank you for being there.

There's going to be with your -- and and I am trying to explain it that you obviously can explain it much better tell us about this new technology.

How it works who invented it and how it's helping it's you it show what people are saying about the Oscar contenders.

Well so computers and software have been great for years -- analyzing numbers and information in columns and rows.

We're what we're doing here is we're actually analyzing human language in the way that it's put forward in.

And in fact twitters and interest you want as you know as it is we all tweet we're not especially focused on know what language for using -- how precise are grammar is so what we're doing -- for using the most.

Advanced analytic software from IBM.

And we -- and in partnership with the Annenberg innovation lab over -- USC.

Are analyzing up to a million tweets a day that are coming out and -- categorizing those for best actor.

Best actress and then picture of the year -- and we're analyzing not only how much activity your buzz there is around each of those categories.

But we're analyzing what the sentiment is as a positive force in negative.

Yeah any keywords tell us what -- -- -- -- is.

We are saying this is this is an ever changing technology considering.

The language that we use on Twitter LO well things like got in what is it really zeroing in on.

And obviously public opinion can change day today.

Whether it be exactly right in and put them part of what happens here is -- first small.

The scale so we can analyze in the Super -- is a good example -- during the super -- we were analyzing.

Up to 121000 tweets the second a day it.

They -- with amber usually we're looking at about a million tweets so this gal is really important and then.

As far as the key -- -- -- -- analyzing the the languishing context the software actually burns is -- goes he gets smarter as he goes.

So over time we get more and more precise about the way were able to measure the sentiment in and to look at what people are actually same name get a sensitive are they seen something that tends to be more positive -- -- negative.

Cell let's break it down we're not predicting Weiner is here here's two people are talking about on winner let's see how some folks are doing so far how is George Clooney for best actor nominee for the descendants doing on the centimeter.

Brought the world as you'd expect George Clooney and and Brad Pitt tend to get a lot of the discussion in -- you have best actor category and both of them.

Tend at least according to the centimeter -- we've been publishing it and it with the LA times and attend he more or less neutral in the best.

Actress category.

Meryl -- tends to get the bulk of the attention in the bulk of the tweets that are that are going out.

And hers can actually able slightly negative on on the -- on what we've been seeing but.

One of the interesting things is because we -- we post this every day.

Things change over time both the amount of discussion it happens.

But also how the sentiment tends to be and if you look on January 15 as an example when the golden gloves.

Were announced there we saw this tremendous growth in the amount of volume on -- there were we're done online.

So what about.

Because that's gotten so much but as you hear it everywhere you go online also is the artist getting just as much buzz.

Yeah and and quite positive actually you know I find the best picture category I'm probably the most interesting of the three were looking at because of the volume of tweets that are coming out so we get more.

To look at but.

But secondly because some of the movements.

The artist is -- tended to be quite positive in the sentiment as we've seen over time and to help as an -- -- -- because.

If we go back has some of that early days -- -- help was first coming out you know it was a relatively smaller picture got a lot of positive momentum.

And I think what we're seeing here is much likely you know when we want to.

Go to a new hotel learn new restaurant and we'll look on the web won't get recommendations -- -- ask and our social network -- our friends are thinking.

I think there's -- momentum aspect to what we see that when something starts to get a lot of positive momentum that builds.

Over the last couple weeks midnight in Paris says that actually -- if you look at the centimeter we've seen it move pretty dramatically in a positive direction so.

So there's movement to -- back and forth over time.

And and the volume tends to grow or shrink over time but those three yet tend to be -- attracting pretty positive.

Interest saying I wonder why -- some of them have new dvd releases coming out and and people are talking about it once is CNN home.

This could be all work in to various marketing strategies in the future you never know.

So thank you know exactly and not just as they think he's I was in the I was -- your comment please please go ahead.

As the city on that point it's you're exactly right we think of this has been applied to and to basically.

Any company in the world is putting products or services and -- marketplace now we have a way.

To analyze in real time.

What consumers are saying about their interaction with your brand which -- company with your products.

-- we can capture all of that information as it's posted on the web at scale and analyze it in real time so.

As we go forward we see this is a really important new way to capture kind of the polls with the public.

And how can people keep the night is -- you'll be monitoring -- The Oscars as soon as they -- You actually absolutely.

That LA times on the website has so what they're calling the centimeter and you can go on in you can actually track day by day.

What the results of our analysis is -- you can see as we -- you know this week I'm sure the the traffic were billed as we approach in the buzz will build as we press the Oscars so you go one on line and take a look at it and you can see how your favorite film or bacteria a is tracking.

Great Steve thank you so much -- really appreciate you being with us today.