понедельник, 10 октября 2011 г.

Datafiniti builds a webscale search engine for data


There’s a lot of talk about the democratization of data, but simply making data sets publicly available leaves unresloved a couple of key problems. One is that users have to know the data is available, and the other is they have to know how to work with it to get the information they want. Austin, Texas-based startup Datafiniti is looking to change that with a search engine that it hopes will make finding relevant data as easy as finding relevant web sites using Google.


Datafiniti’s web-based service is as simple as entering SQL parameters into a search box, clicking “search” and watching the results display. The interface even suggests relevant search terms as users type, based on the categories of data available in the Datafiniti platform and the different fields by which it’s sorted. Once users have the information, they can use the supplied API to feed their application with the information, or they can export it as CSV files or JSON objects.



The service is pretty clearly targeted at developers wanting to build data-centric web applications and, as founder and CEO Shion Deysarkar explained to me, Datafiniti tries to differentiate itself with the aforementioned API. Users get just one API that offers up access to multiple data types, which contrasts with the traditional mashup experience of using a different API for each data source.


As of a couple weeks ago, Deysarkar told me Datafiniti consisted of about 15 million records, but it’s constantly growing and his goal is to index “any and all structured data on the web.” Datafiniti launched today with five available data types — Location, Social or Identity, Product, News and Real Estate — but Deysarkar said there will be many more to come.


Jeff Ferries, Datafiniti’s VP of sales, told me that Datafiniti also targets business users. Firms such as Dun & Bradstreet, Axciom and Experian, he explained, are limited in the types of data they can compile, and they usually can only add data monthly as their source material is updated. Deysarkar said that process, which involves opening a physical catalog, flipping through listings and calling a broker, is “pretty ridiculous,” especially considering that data might be out of date by the time it’s complete.


If Deyasarkar’s name sounds familiar, that’s because he’s the founder of 80legs, a high-powered web-crawling service whose launch my colleague Stacey Higginbotham covered in 2009. 80legs as a company no longer exists, but the service powers Datafiniti by constantly crawling the web and adding new data, and it’s still available as a product from new parent company Datafiniti.


The image above shows a search string for restaurants located in Austin and its results, but this video offers a more-thorough walkthrough of the Datafiniti experience.



Good Technology gives enterprise apps a secure launchpad


Good Technology has already enabled companies to extend security and management to a fleet of mobile devices that consumers are bringing to their IT managers. However, the underlying structure has worked primarily with some basic applications such as email, contacts and calendar.


Now, the company is looking to extend its technology to independent software developers, corporate customers and systems integrators, allowing them to build apps that can ride on Good Technology’s secure platform. The new product, called Good Dynamics, will allow developers to create and manage secure mobile apps using Good’s infrastructure, which acts as a container around the application and all of its data. That should enable companies that have been itching to create their own in-house apps or software developers looking to sell apps to corporations to move ahead with apps that will meet tough security standards, which is important for many business customers.


Nicko van Someren, CTO of Good Technology, told me in an interview that many enterprise customers have worked up apps in-house but they can’t deploy them because they haven’t passed the necessary security requirements. He said with Good Dynamics, enterprise customers can move ahead with those apps while leaving the task of encryption, authentication and connectivity to Good.


“It’s not just about encryption, transport and security, it’s a holistic solution to manage these apps,” van Someren said. “In regulated industries, you need to be seen as keeping data secure.”


Developers can plug Good Dynamics’ code libraries into their apps to secure their data and get access to it. Good Technology maintains the Network Operating Center infrastructure and, in the event of a security breach, customers can kill the Good Dynamics-enabled app without affecting any personal data on the device. Good has already been testing Good Dynamics with a number of software makers and clients, including Aji, Accellion, Box, GroupLogic, MeLLmo, Pyxis Mobile, Quickoffice and Unisys.


Good Technology will offer Good Dynamics alongside its Good for Enterprise, Good forGovernment, and Good for OEM/Carriers products. It should be available for customers next month.


Good Dynamics is a logical next step for Good Technology. The company can try to create apps itself for enterprise customers but it makes more sense to extend its platform to a host of apps already waiting for more security. In this app-addicted world, Good’s in a strong place to capitalize off the habits of companies and end-users.


Warner Bros. closes HBO window ahead of UltraViolet launch


Warner Bros. has confirmed it will no longer be beholden to rights that would see its movies disappear from streaming sites while those titles are available on HBO and its HBO Go web and mobile applications. The closing of this so-called “HBO window” could go a long way toward making digital ownership of UltraViolet titles more appealing to consumers.


UltraViolet hopes to make digital ownership of movies more attractive, by allowing consumers to buy a title once and access it anywhere or on any device. One of the big questions revolving around the impending launch of UltraViolet streaming video services was whether or not studios would have to deal with the rights window, during which HBO has exclusive access to those titles online. Until recently, that meant movies purchased online couldn’t be accessed while HBO had pay TV rights to that content.


On a press briefing Monday afternoon ahead of the launch of the first UltraViolet-enabled title, Horrible Bosses, Warner Home Entertainment execs said that its movies won’t be subject to the HBO window. As a result, anyone who purchases a Warner Bros. DVD or Blu-ray disc won’t have to worry about losing access to the movie online or being blacked out once HBO gets ahold of it.


That doesn’t mean that HBO’s other studio partners — 20th Century Fox and Universal Pictures — aren’t still subject to the same restrictions. Warner Bros. execs wouldn’t comment on where its competitors stood with regards to whether or not HBO had streaming exclusive rights to their movies when they enter the pay TV window. Of course, Warner Bros. and HBO are both part of media conglomerate Time Warner, so their interests are more likely aligned than HBO would be with other studios.


The news that HBO has restructured its deal to allow Warner customers to stream its movies whenever comes at the same time that it is making a big digital push of its own. The premium cable network is making all its movies and original TV series available online, on mobile devices and on connected TV platforms through its HBO Go initiative. Part of Time Warner’s broader TV Everywhere push, HBO Go gives users the ability to access on-demand content as long as they prove that they’re cable subscribers.


Why Silk won’t be silky smooth for Amazon





Amazon in late September launched new Kindle devices including Kindle Fire, a tablet that makes content a centerpiece of its tablet strategy. It also announced a new browser, Amazon Silk, that proposed to use cloud to offer a blazing fast experience. Silk’s hybrid browser architecture quickly triggered some privacy concerns. Amazon weighed in on my queries and clarified their position.


Nevertheless, I have continued to receive feedback, some private and some over various social networks. One that stands out is from Mathew Prince, co-founder and CEO of Cloudflare, a hosted proxy service provider based in San Diego who shared his thoughts. Prince, (you can follow him on Twitter @eastdakota) who teaches cyber law at John Marshall where he serves on the Board of the Center for Information Technology & Privacy Law, believes Amazon will continue to face “technical, legal, and privacy concerns with Silk.” He points out that similar attempts in the past have not been very successful, even for Google.


Amazon’s Silk Browser may be a game changer, but the history of similar efforts shows the company may face significant headaches in getting it to work. The Silk Browser loads pages through a proxy which can have a number of benefits to end users. Depending on how aggressive the Silk proxy is, it could speed up browser performance, allow Kindle devices to get away with slower, less expensive processors, and potentially even increase the battery life by offloading web rendering.


The Silk Browser it isn’t really new technology and it’s not a slam dunk that it will work. The Opera Mini browser uses a proxy which has several of the same features as Amazon’s Silk. Google tried something similar back in 2005 with their Web Accelerator Plugin. While the plugin is no longer available, the support documents still are. Google discontinued support in early 2008 after a number of issues arose — similar issues that are likely to be faced by Amazon with Silk.


I predict that Amazon is likely to face technical, legal, and privacy concerns with Silk. Technically, the biggest challenge will likely be cache invalidation. If I visit my bank website and my account page is cached, Amazon needs to be 100 percent certain that when someone else visits the same bank they never see my account information. From the technical specifications, it appears that Amazon is only caching static resources such as images. While that will solve many of the cases, there will still be places that Silk could end up leaking private data (e.g., a stock photo or porn site that charges for access to its photos).


Unlike existing proxies (like CloudFlare) or traditional CDNs whose clients are the website owners, Amazon’s clients are the web browsers, so they are copying content without the content owners’ explicit permission. This could lead to copyright headaches. While there are safe harbors for service providers caching content, Amazon’s nebulous status between network provider, retailer, and even publisher could muddle their case in court and make them a tempting target. The more Amazon alters the content in order to increase performance, the more jeopardy they will put themselves in.


Finally, Silk potentially puts Amazon in the privacy crosshairs. It appears they are planning to subsidize some of the Kindle’s pricing with advertising, and that advertising will likely be most effective if it is targeted using browsing data gleaned from Silk. Users and regulators can react very strongly if they feel their information is being sold without their permission, and Silk has the potential to score high on the creepiness factor. These privacy concerns have a way of blowing up unexpectedly with regulators resulting in substantially burdensome regulation. In this case, Amazon has already made many government enemies as they’ve fought Internet sales tax initiatives. Going after them for privacy violations may prove a tempting target for lobbyists that already trying to demonize them.


My hunch is that Amazon will find a way to pull it off, but it won’t entirely be smooth for Silk.


What do you think about Prince’s take?


For mobile in-app sharing, Twitter tops Facebook 3-1


Mobile developers are increasingly finding the power of social integration in their apps, which is one way to make them engaging and sticky and to help get the word out about their apps. But all social integrations are not equal.


User by user, Twitter integrations on mobile apps drive three times as much sharing as Facebook integrations, according to new data from app analytics firm Localytics. And the advantage is even more pronounced when you look at active users.


Localytics examined all mobile apps with more than 500 monthly active users connecting to Facebook and Twitter from January through July 2011. It found that 20 percent of all mobile apps on Android, iPhone, iPad, BlackBerry and Windows Phone 7 apps have social integration through Facebook or Twitter. Facebook is the most popular social network by far: 10 percent of all apps connected to Facebook only, compared to just one percent of apps that connect to Twitter only. Nine percent of apps connect to both Twitter and Facebook.


Compared to Twitter, Facebook overall generated twice as many events, which Localytics counts as sharing, liking or following by a person from an app. But on a pound for pound basis, Twitter won out handily when it came to driving user engagement. The average Twitter user shared three times as many events than the average Facebook user, Localytics found. When you examine the active user base of each network, Twitter generated 50 events per 1,000 users compared to 11 events per 1,000 Facebook users, said Localytics.


Now, Facebook has a far bigger community with 800 million users, including 350 million who access the service through a mobile device. Twitter has more than 200 million accounts registered with 100 million monthly active users. So with that size advantage, it’s not surprising that Facebook produces more user events overall.


Daniel Ruby, director of online marketing for Localytics, said it might be that Twitter’s more flexible approach to identity, which allows people to be more anonymous, can prompt more sharing. It’s also probably due to the inherent differences in the two services. Facebook is more of a personal network of friends, while Twitter is often used as a broadcast communications platform for sharing information. I know I find it easier to Tweet something than sharing it on Facebook, where I might be sensitive to overwhelming my friends. But having two options can be helpful for people who might share differently according to each situation.


For developers, it’s important to know what tools are available and how they might differ. Twitter has smaller reach, but it has a very active user base. And with its iOS 5 integration, it might be in a position to grow rapidly. For the four of every five apps that don’t have social integration, their developers might want to think about adding it. As I wrote about earlier, there are a lot of apps that don’t connect to backend services of any kind. But the payoff for those that do are more engaging and dynamic apps, especially ones that have social integration. Those apps can not only keep people involved, but they can draw in new users who can be attracted through sharing on social networks.


And as Localytics points out, developers can also get interesting insight into what features or content are popular on an app based on what people are sharing. That can help them tweak or tune the app accordingly.


Memo to newspapers: Let your readers inside the wall



The Guardian, the U.K. newspaper that has been one of the biggest mainstream-media champions of a “digital first” approach and a proponent of “crowdsourcing” the news, says it’s now going to experiment with allowing readers to help decide what news to cover. The paper announced Monday that it’s going to make its “newslist” — the daily schedule of stories the media outlet thinks are worth covering — public, something which the paper has previously kept carefully guarded. But the Guardian seems to have realized what many newspapers have not: If you allow your readers to be part of the news-creation process, they will be more engaged.


In a column announcing the experiment, Guardian National News Editor Dan Roberts says the idea of showing readers — and everyone else — what stories the paper is working on might seem a little strange, since many publishers try as hard as they can to keep this kind of information secret (in some cases, rival newspapers in Britain have paid leakers for access to a competitor’s newslist). But the Guardian editor says that the paper believes opening up to its readers will improve the quality of its reporting, and help it concentrate on the stories that will be of the most interest and/or value. Said Roberts:


What if readers were able to help newsdesks work out which stories were worth investing precious reporting resources in? What if all those experts who delight in telling us what’s wrong with our stories after they’ve been published could be enlisted into giving us more clues beforehand? What if the process of working out what to investigate actually becomes part of the news itself?


Let readers in and they will help make your work better


Roberts notes that The Guardian isn’t opening up its entire newslist, and will be excluding any exclusives the paper might have, as well as embargoed stories that have been planned in advance. And the editor also says the newspaper is looking at this as an experiment, and is prepared to “pull the plug” on the trial period if competitors are benefiting too much — or if readers simply aren’t interested. That said, however, Roberts says the paper is convinced that offering this information to readers will help focus its news-gathering better and provide tips that can improve stories.


[W]e think there are lots of routine things that we list every day which might provoke interesting responses from readers: everything from upcoming press conferences, to stories we need help uncovering. If readers can see that we’ve got a reporter looking into the police killing of someone with a Taser – to use a recent example – they might be able to direct us to other recent deaths or the definitive report on their safety risks.


What the Guardian is doing is quite simple: it is letting its readers behind the wall, pulling back the curtain to show them some of the machinery involved in producing the news, and offering them the chance to help — a smart approach that other media outlets could and should emulate. It’s an extension of what the British paper did with its groundbreaking “MP Expenses” project in 2009, which involved uploading more than 200,000 official expense reports for British politicians and then asking readers to comb through them looking for errors or fraud. In the end, more than 20,000 people did just that, in one of the most successful crowdsourcing projects ever undertaken by a newspaper.


Using readers as a resource is one thing, but revealing what stories are planned and offering to let readers affect that process is another. In the not-too-distant past, most newspapers were almost as secretive as government agencies; the processes involved in producing journalism day-to-day were only revealed to members of the priesthood, and things like story lists were kept under virtual lock and key. Needless to say, this kind of culture isn’t conducive to things like blogs or story comments or Twitter, as is obvious from many mainstream media companies’ restrictive social-media policies.


They are the “people formerly known as the audience”



But opening up to readers — or what journalism professor Jay Rosen has called “the people formerly known as the audience” — has obvious benefits, as Roberts notes. For one thing, if they’re involved in a particular story, they can provide details and perspectives that might never have come to light during the traditional reporting of a news event. No matter how expert a journalist might be at covering his or her beat, there’s always going to be someone who knows more about that topic, and giving them the chance to contribute to a story makes the end product better, whether reporters want to admit it or not.


As Roberts notes, this approach has paid dividends for other newspapers that have tried it, including a Swedish regional newspaper that has been experimenting with an “open newsroom.” The large daily uses a live-blog powered by CoverItLive (owned by Demand Media) that’s run by a senior editor, which provides a place for the paper’s staff to talk about the stories they are working on, and allows readers to post their comments and questions. The paper’s editor-in-chief says that doing this has not only driven traffic to the site, but created a more engaged readership.


Other newspapers experimenting with open newsrooms include the Register-Citizen in Connecticut, part of the Journal-Register Co. group — which, like The Guardian, has taken a “digital first” approach under CEO John Paton, now the CEO of the Media News Group — and the Winnipeg Free Press in Canada. The Register-Citizen allows readers to come and go as they please in the paper’s newsroom, and provides coffee and Internet access, as well as inviting readers to be involved in story meetings. The Free Press, a large daily, has stationed several reporters in a coffee shop/restaurant in the city’s downtown core, so that readers can interact with the paper’s staff more easily.


The bottom line is that the secrecy that newspapers used to operate under no longer works. Not only is there more competition for stories, but mainstream media outlets no longer have a monopoly relationship with their readers, who can find the same information from dozens of alternative sources. Either newspapers develop a more balanced relationship with the people formerly known as the audience, by allowing them to contribute to the process, or they will find their audience has gone elsewhere.


Post and thumbnail photos courtesy of Flickr users Jeremy Mates and Yan Arief Purwanto


5 trends that will shape the future of mobile advertising


Spending in the mobile advertising space will be approximately $4 billion worldwide this year, so despite some perceptions to the contrary, we can safely assume it’s an area to watch in the coming years.


To get a clearer picture of mobile advertising’s future, it helps to first explore the current landscape. Five trends in particular stand out:


More-relevant behavioral targeting. Currently, mobile advertisers use traditional methods of ad targeting: device, demographic group or context. But in five years, mobile ad targeting will become more relevant to a person’s behavior and current location. For instance, online advertising network Adverline has an AdNext server that builds a prediction model of spatial and temporal relevance for delivering ads to mobile users. The company has tested this with the COEX Mall in South Korea, and it found a 70 percent confidence level that ads displayed are spatially and temporally relevant.


Growing mobile search. Mobile search advertising is already a big driver of ad spending; this is where Google currently makes much of its money in mobile. In five years, this will still be the case, as people continue to use mobile devices to search for products and services. What will change is that, coupled with better targeting, search advertisers will deliver more relevant ads based not only on location and relevant keywords but also on more accurate predictability engines.


Better analytics. Right now, the tools for measuring mobile ad campaigns lack the sophistication that advertisers expect. The Mobile Marketing Association and other groups, such as the Interactive Advertising Bureau, have set up good guidelines. But these are still evolving, and not everyone plays by the same rules. In five years, the standards will have matured. Companies will agree more about what gets measured, and analyzing mobile campaign effectiveness will not be as big of a hurdle.


Greater interactivity. Most mobile ads today are delivered as relatively simple text or banner formats. They are clickable but not very interactive. Rich-media ads are available, too, but they are not as widely deployed for a variety of reasons: more cost, and not all devices render them well. In five years, interactive ads will be common, and users will be expanding, collapsing and manipulating them in ways that are still unfolding. This will unleash creative minds to show off their capabilities. Also, technology like augmented reality will make the interaction with brands and the world around consumers more interactive: Imagine looking at your mobile screen and playing a Zynga game sponsored by Coca-Cola, based on your physical surroundings.


Mobile-social as the “personal cloud.” Today the mobile and social worlds already collide with services like Foursquare and Gowalla. In five years, though, this functionality will likely evolve into what some Stanford researchers suggest is the personal cloud, which will be personal data that surrounds us wherever we go. It will also be shareable with whomever we choose and will be used to make purchases, among other things. Savvy advertisers will leverage this trend with relevant offers that provide value but still respect user privacy.


Meanwhile, companies like Google, Apple and Millennial Media currently rule the mobile advertising space, and while some of those will continue to have an impact over the next five years, others leading the way today will fade away or be acquired. To read about these, as well as more trends to watch, please see “The future of mobile advertising, 2011-2016” at GigaOM Pro (subscription required).


Image courtesy of Foursquare.