Every day, Yahoo displays about 13 million different news story combination on its homepage. Those stories are personalized based on demographic data and reading behavior, and the company keeps track of what kind of stories do well with which groups of people.

Sponsor

To do that, Yahoo utilizes a complex set of algorithms it calls the Content Optimization and Relevance Engine (CORE). The system crunches 1.2 terabytes of data per hour to determine which stories to deliver to which users. The result is a line-up of stories on the homepage that’s customized for each user, based on calculations that take milliseconds to crunch as the page loads. It also lead to a substantial increase in engagement on Yahoo’s site, where click-throughs to news stories have increased by 300% since this technology was first implemented.

To illustrate how this works, Yahoo has created an interactive data visualization that shows visitor traffic data in nearly real time. Using it, one can drill down into specific age groups, genders and story types to see what people’s aggregate reading habits look like.

You can view and play with the data visualization here. They even designed the UI in HTML5 rather than Flash so you can check it out on your iPad.

Discuss

View full post on ReadWriteWeb

Every day, Yahoo displays about 13 million different news story combination on its homepage. Those stories are personalized based on demographic data and reading behavior, and the company keeps track of what kind of stories do well with which groups of people.

Sponsor

To do that, Yahoo utilizes a complex set of algorithms it calls the Content Optimization and Relevance Engine (CORE). The system crunches 1.2 terabytes of data per hour to determine which stories to deliver to which users. The result is a line-up of stories on the homepage that’s customized for each user, based on calculations that take milliseconds to crunch as the page loads. It’s also lead to a substantial increase in engagement on Yahoo’s engagement, where click-throughs to news stories have increased by 300% since this technology was first implemented.

To illustrate how this works, Yahoo has created an interactive data visualization that shows visitor traffic data in nearly real time. Using it, one can drill down into specific age groups, genders and story types to see what people’s aggregate reading habits look like.

You can view and play with the data visualization here. They even designed the UI in HTML5 rather than Flash so you can check it out on your iPad.

Discuss

View full post on ReadWriteWeb

View full post on Search Engine Watch – Latest

View full post on Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Sponsor

While many in tech journalism circles have been quick to call the 2012 presidential race “the Social Media Election,” the poll found that few of us are relying on Facebook, YouTube and Twitter for election information. While 25% say they regularly learn something about the election from the Internet, tha’s almost unchanged from 2008, when 24% said they regularly got election information from the Internet.

Even more telling is where on the Internet that information comes from: 6% of poll respondents said they are regularly learning about the campaign from Facebook, followed by YouTube videos (3%) and Twitter (2%), according to Pew.

One reason social media hasn’t grown by the leaps and bounds predicted is less engagement by young people. In 2008, there were two contested primaries, including a Democratic primary which has traditional drawn younger and arguably more tech-savvy voters. This year, only one in five people under 30 say they have been following the campaigns “very closely,” down from 31% in 2008.

Discuss

View full post on ReadWriteWeb

Donna Burbank – who’s a senior director of product marketing for CA Technologies’ long-time data visualization tool, ERwin, has a new phrase for this class of customers: business sponsors. “When I talk to our customers, they tell me it’s a whole new… thing, for lack of a more technical word. They’ve heard of SQL Server, but what is this SQL Azure thing? They don’t have the skill sets, and may be nervous about that. These business sponsors might not be moving the information, but they want to see it. And they don’t want to look at those database scripts. They want to look at something they can understand.”

Sponsor

So it is that CA Technologies found itself in the business of manufacturing a class of software that a new and growing chunk of its customers might not actually care about all that much: database visualization tools. ERwin has been the market share leader in this category ever since its creation in 1998. Its typical customers have been database architects (DBAs), the people whose jobs are to model the classifications and structures for the relational data that businesses rely upon every day.

Welcome to Your World

But the shift to cloud technologies has partly been fueled by the need for tremendous space for data warehousing – to house the huge data stores generated by millions of Internet customer transactions. It’s that shift which is pushing data outside of the constraints of traditional SQL relational databases. That push is forcing businesses to examine, some for the very first time, the structure of their data. And what they’re seeing, they don’t understand.

“A lot of the move to the cloud is a business decision. The technical people doing the move are probably still the DBAs, but they’re challenged,” CA’s Burbank tells RWW. “There will always be that core group of people who want to use a data model, that’s our sweet spot right now: the data architect, the DBA. Those people and more would like to use a Web-based interface.”

ERwin’s new Web portal, she explains, is a browser-based interface for information that has otherwise been modeled for DBAs by ERwin Data Modeler (which itself moves to version 8.2 this week). This new portal will help both architects and Burbank’s “sponsors” to analyze the relationships of data from a business impact standpoint. “If I’m building a data warehouse, I want to see how data moves from the source system to the target warehouse to the reporting tool. Maybe I’m changing a data element; what other parts of the organization are affected? You could sort of get that through ERwin’s repository [in Model Manager] with some queries, but it wasn’t the tool for that.”

The new, more objective breakdown aims to give multiple classes of users a comprehension of the data that they may have never had before. A “sponsor” who wants to search for relationships is going to expect search to behave like Google, Burbank explains. So the Web Portal tool gives that user a text-based search query line (shown above). What that user gets in return will be something that may explain what tables or fields relate to the search criteria, but it might not directly correlate to the model as the original DBA intended.

As Burbank explains, the business user, not getting a complete overview from the initial response, may decide to export the data he’s seeing into Excel, and generate some PivotTables from them. The DBA, on the other hand, may use the Portal’s new graphical impact analysis tools to drill down further, or perhaps execute a “What If?” experiment. If a column is changed on a table, for instance, the DBA can see how the rest of the schema is impacted. “It’s that type of drilldown over the Web that they could never do before,” she remarks.

Big data vs. “lots of data”

All this said, ERwin is not quite yet a data warehousing assistance tool. While CA’s Donna Burbank says it’s something her company is considering, she points out that Hadoop and the restructuring of data it entails, lend themselves to very different situations.

“I’ve done several presentations where I’ve explained to people that there’s ‘big data,’ and then there’s ‘lots of data,’” she relates. “And these are different use cases. Maybe I’m an energy company, and I’m trying to use a Hadoop-type structure to see uses across my different [operating units], and I need to eventually manage that in a warehouse. It’s that analysis of that big data that then goes into a data model. One use case [involves] massive volume, real-time, more of a programmatic approach to data. There’s a lot of messaging there around, is data modeling going away? Is data warehousing going away? Today, it’s two different use cases. You’re doing an analysis, and then you use the data model to make sense of that raw data. And if I’m going to use it for a BI report, that’s when your data model comes in. I’ve done my data analysis with the big data; here’s my data model to say which pieces of that I used in the warehouse.”

Real-time analysis of big data, she goes on, may enable DBAs to add some elements to the relational data model that they may not have seen before.

As for the other use case, Burbank agrees that data modeling may never be appealing to 100% of the “sponsor” audience. But making it appeal to a somewhat greater audience through more intuitive graphics, along with Google-like search, could go a long way toward enabling those tasked with new responsibilities to be able to better understand what they are, and carry them out with a greater sense of confidence.

Discuss

View full post on ReadWriteWeb

View full post on Search Engine Watch – Latest

View full post on Search Engine Watch – Latest

More than 3 million people are now paying to use Spotify, according to the Financial Times. That’s a conversion rate of more than 20%, a figure that has reportedly increased by 5% since the service hit 1 million users last year. In other words, not only is Spotify itself growing, but the rate at which people sign up for a premium or unlimited account is also increasing.

Sponsor

This overall growth has been fueled in no small part by the company’s partnership with Facebook, which enables the kind of super-tight, frictionless integration that the social networking giant has been pushing since f8 last March. The flood of “so-and-so listened to such-and-such” news ticker updates may be too much for some people, but the partnership has succeeded in putting Spotify’s brand and functionality in front of millions of potential new users.

It also doesn’t hurt that the six-month window of unlimited, free streaming music for new users has begun coming to a close for the service’s earliest U.S. adopters. As that happens, those who are truly hooked on the service are forced to either put up with listening caps or cough up $5 per month to remove them. The company hasn’t said what percentage of those paid users have opted for the pricier “Premium” account, which allows for mobile streaming in addition to stripping out ads and listening caps.

This isn’t to say that there aren’t still major, outstanding questions about Spotify and the viability of the all-you-can-stream model it shares with the likes of Rdio and MOG. The music labels are evidently happy enough with the arrangement to stay on board for now, but the artists are a different story.

The streaming services pay out notoriously low royalty fees to artists, some of which have begun to question the value in being on the service. Sure, it’s a great way to promote one’s music, but it may not be economically advantageous for artists, especially if it ends up hurting record sales.

Discuss

View full post on ReadWriteWeb

Nearly 1 million people jailbroke their iPhone 4S or iPad 2 between Friday and Monday, according a blog post from the Chronic-Dev Team, who took the lead in developing the untethered solution for jailbreaking iOS 5 on Apple’s newest gadgets.

Sponsor

News of Friday’s launch of Absinthe A5 temporarily crashed the greenpois0n site, as it evidently gave hundreds of thousands of users a fun weekend mini-project. The initial tool was only released for Mac OS X, but a Windows version quickly followed over the weekend.

Not surprisingly, the iPhone 4S was the most-jailbroken device with over 491,325 phones broken free from the restrictions of the iTunes App Store. Since iOS 5 came pre-installed on the iPhone 4S, this is the first time its owners could jailbreak the device without tethering it to a computer, which is too cumbersome a process to be considered by most users. The iPad 2 saw n 308,967 new jailbreaks, on top of the 152,940 second generation iPads running iOS 4 that were re-jailbroken. Those users were likely waiting to upgrade to iOS 5 until this solution was available, which is one of the drawbacks of jailbreaking.

Why People Jailbreak

For devotees of the jailbreaking process, the drawbacks are vastly overshadowed by the benefits. Rather than be able to cite a single “killer app” that makes it worth it, most just prefer the general freedom and customizability it offers. That includes the ability to tweak the visual appearance of the device’s UI and run any number of unauthorized apps.

Cydia, the repository of apps for jailbroken iPhones and iPads, contains many applications that would never meet Apple’s approval requirements for inclusion in the official App Store. For some, it’s because of trademark or copyright issues, such as video game emulators or controversial music services like Grooveshark.

Quite often, the apps don’t adhere to Apple’s agreements with the carriers, who would obviously never approve of an app that lets users tether their phone to their laptop and use its data connection without paying extra. With the iPhone 4S, the feature that offers perhaps the most new potential for jailbreakers is Siri, which developers have wasted no time hacking and tinkering with.

As controversial and officially frowned-upon as the whole thing is, many of the experimental features available on jailbroken iOS devices actually end up being included in the next release of the OS. This was true of recording video, which was technically possible on a jailbroken iPhone 3G. Similarly, the overhauled Notification Center found in iOS 5 bears a striking resemblance to the notification system available on jailbroken devices running iOS 4.

Jailbreaking May Not Be Mainstream, But Its Popularity is Growing

As Apple’s gadgets continue to burn through sales records, there also appears to be a growing interest among users in doing more with their phones, much like Android device owners are already accustomed to.

Truth be told, jailbreaking probably isn’t something the average user is going to get into, given the nature of the process, its risks and the tools involved. It’s really something more geared toward tinkerers and those who want more control over what their device can do.

Given the massive number of iOS devices out in the wild, several hundred thousand new jailbreaks doesn’t come anywhere close to a constituting majority of of users. Rather, it’s the pace of the growth that’s interesting to see.

Discuss

View full post on ReadWriteWeb