Wednesday, April 7, 2010

Googland

Googland


[G] Curator of the Month: Ad Age

Posted: 06 Apr 2010 09:53 PM PDT

YouTube Blog: Curator of the Month: Ad Age

We're pleased to have Ad Age curating the YouTube homepage today, in a spotlight that celebrates the creative revolution going on now in the advertising and marketing business. In other words, these are ads you won't want to miss.  Ad Age digital lead Michael Learmonth and Creativity managing editor Ann Diaz explain further in the video and guest post below:






















We, in the business press, love to obsess over YouTube's business model, whether it can make money from the world's video through advertising. But today Advertising Age is curating the front page of YouTube to help tell an even bigger story, and that's YouTube's impact on advertising itself.



Once, TV ads were pretty much foisted on the public. Turn on the TV, and they were there. Some were great; most were not. Indeed, in some TV ads the intent is to annoy and grab the attention of a passive public. Enter YouTube. And while all of that is still true, the marketing world now has another powerful, democratic vehicle to reach a TV-sized audience. But there's a catch: the ad has to be something people want to watch.


Each day, YouTube is a global referendum on the world's video, and ads are very much a part of that mix. TV ads have always had the power of sight, sound and motion; now, to reach an audience in an on-demand world, they also have to delight, entertain and tell a story. That has inspired a creative revolution in the advertising and marketing business, just as it has in entertainment and attracted new talent to the industry. It has also refocused the industry away from obsessing over who's skipping the ads to producing ads no one wants to skip. Just like you can buy a 30-second spot on TV, you can also buy media on YouTube, but you can also earn an audience there, and increasingly ads, both made for the web and for TV, attempt to do just that.



Witness the Super Bowl advertisers, on the hook to the tune of $100,000 a second for time in the big game, increasingly gear their campaigns to live on on YouTube after its over. Or, take Evian's "Live Young" aka the "rollerskating babies," which never appeared on TV but have been passed around and watched more than 71 million times over the past year. The beauty of what works on the web is that there are no hard and fast rules. Ad Age picked some of the best of the recent best with a big hat tip to our sister pub, Creativity. It's true that YouTube sells ads. But it's also true YouTube has made advertising better. Take a look.



















Michael Learmonth, Advertising Age, and Ann Diaz, Creativity


URL: http://feedproxy.google.com/~r/youtube/PKJx/~3/tEIjdWg6SMY/curator-of-month-ad-age.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] PyPy 1.2 Released

Posted: 06 Apr 2010 04:27 PM PDT

Google Open Source Blog: PyPy 1.2 Released

PyPy is a reimplementation of Python in Python, using advanced techniques to try to attain better performance than CPython. Many years of hard work have finally paid off. Our speed results often beat CPython, ranging from being slightly slower, to speedups of up to 2x on real application code, to speedups of up to 10x on small benchmarks. This post describes what we did on PyPy during the last year, leading up to those results.

In the spring of 2009 we completed an update of PyPy to support
Python version 2.5 with much appreciated financial support from Google. Most of the work on the update was done by Maciej Fijalkowski and Samuele Pedroni. While this work was in progress, Armin Rigo and Carl Friedrich Bolz were hard at work rebuilding the framework of the Just In Time compiler (JIT). The old framework, using techniques based on Partial Evaluation, only gave good results in constrained cases, but it would usually generate far too much code for Python. It was time to do more research from scratch. What we discovered was that the techniques typical of Tracing JIT compilers better suit the optimization of dynamic languages than techniques based on Partial Evaluation. However, we still follow our original meta-programming approach and remain convinced that writing a JIT compiler generator is more promising than the typical hand-coding of a JIT compiler for a particular interpreter. In other words, like in the original attempt, we get a JIT compiler that is not tied to Python but is generated from the source code of any interpreter for any dynamic language. If you are interested in the details, there is a very approachable paper about it, along with various blog posts.

During the autumn we applied and refined the JIT framework, added more optimisations and wrote a code generator for the x86 CPU family. In the early stages we could get good speed only at the price of huge memory consumption, but much work was spent addressing this problem. We have now reached a point where memory consumption is usually reasonable, but it is the nature of JIT compilers to trade some memory for speed. We began work using the benchmarks from The Great Computer Language Benchmarks Game to identify some problem areas. We have another blog post about our work in this area for the curious. Thanks go to Andrew Mahone who ported many of the Alioth benchmarks for our use. We also did some work on benchmarks that behave more like actual applications. The Django templating engine is twice as fast, and Twisted benchmarks are up to 2.85 as fast. For details, see our progress reports from the blog. While the results on a substantial number of benchmarks are really good, there is a lot more to do. We still have spots where performance is fairly bad, for instance our regular expression engine and the handling of generators. We have ideas about how to improve this and we have a list of further optimisations that could be performed.

The largest issue preventing users from adopting PyPy is the lack of extension modules. In addition to his constant efforts in making sure that PyPy runs on the Windows platform, Amaury Forgeot d'Arc has managed to port Oracle bindings to PyPy, while Alexander Schremmer has worked out a way to use the Remote Procedure Call library RPyC to use CPython extension modules with PyPy. Alexanders' goal was to get PyQt to run with PyPy and he was quite successful (apart from some bugs with PyQt itself), which you can read about on our blog. We also had Benjamin Peterson single-handledly rewrite our previously slow and problem-ridden Python parser, which now is much leaner and meaner. It is beta software, in the sense that it may speed up your applications significantly or not at all. We will need your help finding the odd quirks that prevent your Python programs from running and help to spot the places where performance can be improved. More information can be found on the PyPy blog, at our website or on the #pypy IRC channel at freenode.net.

By Bea Düring, PyPy Team
URL: http://google-opensource.blogspot.com/2010/04/pypy-12-released.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] Marlee Matlin’s new YouTube show and visit to Google

Posted: 06 Apr 2010 04:27 PM PDT

YouTube Blog: Marlee Matlin's new YouTube show and visit to Google

"The real handicap of deafness lies not in the ear; it lies in the mind." - Marlee Matlin, March 30 talk at Google



I never expected that working at Google would allow me to meet a cherished idol, Marlee Matlin. Marlee's acting prowess couldn't help me with my work as a software engineer, yet we faced a common challenge: being deaf in a hearing world. Like many other people, I was inspired by the way she succeeded on her own terms.



That's why it was such an honor for me personally to have Marlee visit Google to discuss online captioning and deliver a talk to employees. As you may know, Marlee is an Academy Award winning actress, author and a national spokesperson for closed captioning access on behalf of the National Association for the Deaf (NAD) and other organizations. She talked about her autobiography, I'll Scream Later, and accessibility issues facing people who are deaf or hard of hearing. In addition, she shared a preview of a pilot for her new reality show on YouTube, "My Deaf Family." We've published Marlee's talk (with captions, of course!) and encourage you to check it out:













Marlee said she chose to publish her show on YouTube, "where I can call the shots and where I can guarantee the show will be broadcast with captions." Since posting online, she's already gotten more than 70,000 views and some great feedback and ratings. People from all around the world are checking out the show and sharing with friends. Take a look at the views rising recently:







As we discussed with Marlee and at our event last November, the online population of people who are deaf or hard of hearing rivals the populations of people speaking many major world languages, such as Italian and Russian. Adding captions to video can help ensure the widest audience possible sees it, and with machine translation, that audience can expand to include people who speak any of 50 languages.



While expanding automatic captions to all users was an important step, we still encourage you to add manual captions to your videos. With our automatic-timing technology it's easy, and manual captions are generally more accurate. We showed Marlee how to add captions to her new show in just a few minutes -- right before she got up on stage for the premiere.



We want to thank Marlee for taking the time to visit us, and for sharing her talent and vision with the broader YouTube community. We wish her the best of luck with her new show, and look forward to working together to expand online access. Personally, I'm still pinching myself after the visit; it's not everyday that I get a compliment from a person like Marlee.



Ken Harrenstien, Software Engineer, recently watched "Jason Molenda In the Net."


URL: http://feedproxy.google.com/~r/youtube/PKJx/~3/nChcrGFyDcM/marlee-matlins-new-youtube-show-and.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] Go Mobile with Google Site Search

Posted: 06 Apr 2010 01:27 PM PDT

Official Google Enterprise Blog: Go Mobile with Google Site Search

As more and more people switch to smart phones as their primary entry point for browsing the web, businesses are looking for ways to help mobile visitors find the right information, fast.

That's why the Google Site Search team has made it easy to enable users to search your website from devices like Android-powered phones, iPhone, iPod Touch, iPad, and Palm Pre.

As a Google Site Search customer, you can repurpose your content to mobilize your web site. You can use the Site Search home page that we create for you as the preferred mobile entry point for your website. All the Google Site Search features - themes, result biasing, promotions, refinement labels, rich snippets, synonym enhancements, etc. – are available on the mobile version as well.

Additionally, if you customize Google Site Search on your website, those features will show up on your mobile home page. If you select or change the theme for your search engine, your mobile home page will automatically pick up those changes. Mobile results will also display thumbnails and actions if you have marked up your pages.

Learn more about these and other features at google.com/sitesearch.

Posted by Anna Bishop, Google Site Search team
URL: http://googleenterprise.blogspot.com/2010/04/go-mobile-with-google-site-search.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] Gmail for mobile integrates with Google Buzz

Posted: 06 Apr 2010 01:27 PM PDT

Official Google Mobile Blog: Gmail for mobile integrates with Google Buzz

When Google Buzz for mobile launched in February, we created the mobile web app at buzz.google.com so you could post buzz on the go. Depending on your inbox integration settings buzz gets sent to your Gmail inbox when people @reply you, comment on something you've posted, or comment on a post after you.

With the latest iteration of Gmail for mobile, we've worked to integrate buzz with your mobile inbox. Now, you'll see buzz in your inbox on your phone just like you do in the desktop version of Gmail, complete with the little Google Buzz icon. When you open a buzz post from your inbox, you can perform all the standard functions, such as liking the post or commenting, just as you can from the desktop Gmail inbox.

To try this out, simply go to gmail.com in your iPhone or Android browser.

Alex Kennberg, Software Engineer, Google Mobile
URL: http://googlemobile.blogspot.com/2010/04/gmail-for-mobile-integrates-with-google.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] Google Buzz Tips for Teens

Posted: 06 Apr 2010 01:27 PM PDT

Google Public Policy Blog: Google Buzz Tips for Teens

Posted by Scott Rubin, Public Policy Sr. Manager

We created Google Buzz to make it easy for everyone* - including teenagers - to start conversations about the things you find interesting. A lot of people have been using Buzz since it launched, but some folks may still have questions about how sharing and other parts of Buzz work.

We made this video to help you and your teens have a conversation about Buzz. It's part of our ongoing series of videos to educate teens about how they can make smart choices online.



You can find more teen safety tips by visiting our Safety YouTube Channel. And for more tips on how to use Google Buzz, check out the Google Buzz YouTube Channel.

*Buzz is for everyone, except those under 13. Like with Gmail, you must be at least 13 years of age to use Google Buzz.
URL: http://googlepublicpolicy.blogspot.com/2010/04/google-buzz-tips-for-teens.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] Collaborative mapping for major disasters

Posted: 06 Apr 2010 10:27 AM PDT

Google LatLong: Collaborative mapping for major disasters

[cross-posted from the Official Google Enterprise Blog]

Editor's Note:
Today's guest author is A.J. Clark, President of Thermopylae Sciences and Technology, which helps developers build new applications with the Google Earth browser plug-in on Google Earth Enterprise through the iSpatial framework. A.J. works on the visualization, analysis, and dissemination of disaster/post-conflict data with a focus on emerging technologies that support participatory, collaborative approaches to spatial content creation and infrastructure development.

In the aftermath of the Haiti earthquake, an urgent need for geospatial data sharing arose. As aid organizations and governments rushed to help, the collection and sharing of large amounts of geographic information among many sources and users presented a daunting challenge.

In response, the United States Southern Command (
USSOUTHCOM) made information and tools available through an interactive mapping website which they call a User Defined Operational Picture (UDOP), built on the Google Earth Enterprise platform to make visualizing the map and overlaid data easy and clear. Public sites are now available both for the Haiti earthquake and the Chile earthquake, where thousands of users can view, create, and edit spatial data (please note that you will be prompted to download the Google Earth plugin to view this data through these links.)




The system was rapidly deployed using an enterprise geospatial framework called iSpatial, which provides an open platform for the integration of dynamic data and the development of interactive applications. The data is displayed using the Google Earth browser plug-in and API, which enables the visualization of geographic data layers in the rich, 3D Google Earth environment, on any web page.

The breadth of data sources and content producers contributing to and making use of the UDOP reflected the global community supporting the relief effort. The combination of participatory geospatial content, the collaborative nature of the UDOP and the ability to integrate mobile applications as direct content producers is available for ongoing relief efforts. This is an evolving capability within SOUTHCOM as we break new ground in providing collaborative tools for humanitarian aid and disaster relief efforts.

Note: to see imagery and maps in Google Earth, you can also download the Google Crisis Response KML files for Haiti and Chile.

Posted by Jeff Martin, Google Crisis Response Team
URL: http://google-latlong.blogspot.com/2010/04/collaborative-mapping-for-major.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] Lessons learned developing a practical large scale machine learning system

Posted: 06 Apr 2010 08:44 AM PDT

Official Google Research Blog: Lessons learned developing a practical large scale machine learning system

Posted by Simon Tong, Google Research

When faced with a hard prediction problem, one possible approach is to attempt to perform statistical miracles on a small training set. If data is abundant then often a more fruitful approach is to design a highly scalable learning system and use several orders of magnitude more training data.

This general notion recurs in many other fields as well. For example, processing large quantities of data helps immensely for information retrieval and machine translation.

Several years ago we began developing a large scale machine learning system, and have been refining it over time. We gave it the codename "Seti" because it searches for signals in a large space. It scales to massive data sets and has become one of the most broadly used classification systems at Google.

After building a few initial prototypes, we quickly settled on a system with the following properties:

  • Binary classification (produces a probability estimate of the class label)

  • Parallelized

  • Scales to process hundreds of billions of instances and beyond

  • Scales to billions of features and beyond

  • Automatically identifies useful combinations of features

  • Accuracy is competitive with state-of-the-art classifiers

  • Reacts to new data within minutes

Seti's accuracy appears to be pretty decent. For example, tests on standard smaller datasets indicate that it is comparable with modern classifiers.

Seti has the flexibility to be used on a broad range of training set sizes and feature sets. These sizes are substantially larger than those typically used in academia (e.g., the largest UCI dataset has 4 million instances). A sample of the data sets used with Seti gives the following statistics:


Training set sizeUnique features
Mean100 Billion1 Billion
Median1 Billion10 Million


A good machine learning system is all about accuracy, right?

In the process of designing Seti we made plenty of mistakes. However, we made some good key decisions as well. Here are a few of the practical lessons that we learned. Some are obvious in hindsight, but we did not necessarily realize their importance at the time.

Lesson: Keep it simple (even at the expense of a little accuracy).

Having good accuracy across a variety of domains is very important, and we were tempted to focus exclusively on this aspect of the algorithm. However, in a practical system there are several other aspects of an algorithm that are equally critical:
  • Ease of use: Teams are more willing to experiment with a machine learning system that is simple to set up and use. Those teams are not necessarily die-hard machine learning experts, and so they do not want to waste much time figuring out how to get a system up a running.

  • System reliability: Teams are much more willing to deploy a reliable machine learning system in a live environment. They want a system that is dependable and unlikely to crash or need constant attention. Early versions of Seti had marginally better accuracy on large data sets, but were complex, stressed the network and GFS architecture considerably, and needed constant babysitting. The number of teams willing to deploy these versions was low.

Seti is typically used in places where a machine learning system will provide a significant improvement in accuracy over the existing system. The gains are usually large enough that most teams do not care about the small differences in accuracy between different flavors of algorithms. And, in practice, the small differences are often washed out by other effects such as better data filtering, adding another useful feature, parameter tuning, etc. Teams much prefer having a stable, scalable and easy-to-use classification system. We found that these other aspects can be the difference between a deployable system and one that gets abandoned.

It is perhaps less academically interesting to design an algorithm that is slightly worse in accuracy, but that has greater ease of use and system reliability. However, in our experience, it is very valuable in practice.


Lesson: Start with a few specific applications in mind.

It was tempting to build a learning system without focusing on any particular application. After all, our goal was to create a large scale system that would be useful on a wide variety of present and future classification tasks. Nevertheless, we decided to focus primarily on a small handful of initial applications. We believe this decision was useful in several ways:

  • We could examine what the small number of domains had in common. By building something that would work for a few domains, it was likely the resulting system would be useful for others.

  • More importantly, it helped us quickly decide what aspects were unnecessary. We noticed that it was surprisingly easy to over-generalize or over-engineer a machine learning system. The domains grounded our project in reality and drove our decision making. Without them, even deciding how broad to make the input file format would have been harder (e.g., is it important to permit binary/categorical/real-valued features? Multiple classes? Fractional labels? Weighted instances?).

  • Working with a few different teams as initial guinea pigs allowed us to learn about common teething problems, and helped us smooth the process of deployment for future teams.

Lesson: Know when to say "no".

We have a hammer, but we don't want to end up with bent screws. Being machine learning practitioners, it was very tempting for us to always recommend using machine learning for a problem. We saw very early on that, despite its many significant benefits, machine learning typically adds complexity, opacity and unpredictability to a system. In reality, simpler techniques are sometimes good enough for the task at hand. And in the long run, the extra effort that would have been spent integrating, maintaining and diagnosing issues with a live machine learning system could be spent on other way of improving the system instead.

Seti is often used in places where there is a good chance of significantly improving predictive accuracy over the incumbent system. And we usually advise teams against trying the system when we believe there is likely to be only a small improvement.


Large-scale machine learning is an important and exciting area of research. It can be applied to many real world problems. We hope that we have given a flavor of the challenges that we face, and some of the practical lessons that we have learned.
URL: http://googleresearch.blogspot.com/2010/04/lessons-learned-developing-practical.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] State Library of Kansas in Topeka goes Google

Posted: 06 Apr 2010 07:26 AM PDT

Official Google Enterprise Blog: State Library of Kansas in Topeka goes Google

Editor's Note: Today's guest blogger is Patti Butcher, Director of Statewide Resource Sharing for the State Library of Kansas in Topeka, Kansas. She received her MLS from Emporia State University in 1993. In addition to her current position she previously worked at the Central KS Library System, Northeast KS Library System, and has served as the State Librarian of South Carolina.

Last week, on April 1, the same day that Google re-named itself "Topeka," the State Library of Kansas (SLK) decided to "go Google" and switch our agency's email to Google Apps. But, unlike Google, we weren't fooling around!

The State Library of Kansas provides on-site and online information services for state government, the Legislature, all Kansas libraries and Kansas residents. We also work closely with seven regional library systems geographically distributed throughout our state. We have a large collection of print and audiobooks, plus state and federal documents. We manage a suite of online resources available 24/7 to all Kansans – from databases and digital books to online tutoring and services for the visually impaired.

Google Apps will provide the library's staff with Gmail, calendar, video chat, real-time document and video sharing, backups, and additional services such as archiving powered by Postini. It will, for the first time, allow the agency's multiple locations to have a staff intranet using Google Sites and join together teams that were previously separated by several firewalls.

The State Library relies heavily on technology so this chance to modernize our communications and collaboration systems is a welcome one. We switched to Google Apps for many reasons:
  • Substantial cost savings
  • Hosting by a trusted third party
  • Archiving and backups happening continuously in the cloud
  • 99.9% uptime, guaranteed
  • Strong recommendations from staff at the Northeast Kansas Library System, which made the switch last year
Now SLK's IT staff will be able to focus on internal staff training needs and upgrading all of our technology systems – rather than on server maintenance. As we deploy Google Apps in the coming weeks, we will first be training our management team who will then work 1-on-1 with all staff to ensure that questions are answered and staff are comfortable with the new products and services. A number of State Library staff already have Gmail accounts and are familiar with its functionality.

Staff members are also interested in Google Apps capabilities beyond Gmail – like Google Groups and Google Docs. Our staff is eager to take advantage of the capabilities of Google Apps!

Staff from SLK will be at the Kansas Library Association conference in Wichita beginning tomorrow, Wednesday, April 7, and encourage anyone interested in Google Apps to seek us out.

Posted by Serena Satyasai, The Google Apps Team

Do you have an informative and fun Google Apps story to share? Please submit it here.
URL: http://googleenterprise.blogspot.com/2010/04/state-library-of-kansas-in-topeka-goes.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

[G] Confirm your Buzz settings

Posted: 05 Apr 2010 10:26 PM PDT

Google Public Policy Blog: Confirm your Buzz settings

Posted by Todd Jackson, Product Manager

(cross-posted from the Official Gmail Blog)

Shortly after launching Google Buzz, we quickly realized we didn't get everything right and moved as fast as possible to improve the Buzz experience. We made a number of changes to the getting started experience based on your feedback, the most significant of which was replacing auto-following with suggestions for people to follow.

Rather than automatically setting you up to follow the people you email and chat with most, Google Buzz now suggests people for you to follow instead. This way, Buzz is still simple to set up (no one wants to peck out an entire social network from scratch) but you aren't set up to follow anyone until you choose to do so.

But many of you started using Google Buzz before we made these changes, and we want to help you ensure that Buzz is set up the way you want. Offering everyone who uses our products transparency and control is very important to us, so if you started using Google Buzz before we changed the start-up experience, you'll see the following confirmation page the next time you click into the Buzz tab:



This page highlights your current Buzz settings and makes it easy to change anything you want. You can view and edit the people you're following and the people following you, elect whether you want those lists appearing on your public Google profile, and modify any of the sites you have connected to Google Buzz, like Picasa, Google Reader, or Twitter. If everything looks good, you can confirm your Buzz set-up with a single click. And remember, you can always change who you're following by clicking "Following XX people" from the Buzz tab or modify your preferences from the Buzz section of Gmail Settings.

To keep up to date with all of the improvements we're making to Google Buzz and provide your feedback, follow our team's Google Buzz account. For tips and tricks on using Google Buzz, check out the videos on our new YouTube channel: youtube.com/googlebuzz.
URL: http://googlepublicpolicy.blogspot.com/2010/04/confirm-your-buzz-settings.html
Feed Ads By BidVertiser.comFeed Ads By BidVertiser.com

No comments:

Post a Comment