Google Data

dinos is a Japanese mail order company that relies on catalogs, TV and the Internet for its business. In their online shopping site, visitors can purchase items they’ve seen on TV or in catalogs. Their site was established in September 2000 and, as of July 2012, dinos now receives 2.5 million user visits and 58 million page views per month. Revenue from their site accounts for about half of dinos’ total revenue.

We recently interviewed Hiroyuki Kikuchi from the dinos marketing team, to find out more about the company’s experiences with AdSense.

■ Inside AdSense (IA): Could you tell us why and how you started to use AdSense?
We started to use AdSense to diversify the sources of revenue from the site and also to monetize the traffic from users leaving the site. The major reason why we decided to use AdSense is because of its flexibility: we can shut off or turn on a campaign at any time. We started out with a pilot in April 2012. Then in May, we began using AdSense more, dramatically increasing the number of pages with ads. As a result, our revenue increased significantly.

■ IA: Did you use any other service before you implemented AdSense?
We used an affiliate service before using AdSense. We piloted this service because it provided control over which ads we served. We also analyzed the results from our pilot to be able to optimize our ad network implementation later on. Our main concern with trying AdSense was that we would not be able to choose the ads we served, but after implementation, we found that we could easily control the ads at the URL and ad category levels. The revenue coming from AdSense is much higher than the revenue we earned from the affiliate network.

■ IA: When implementing AdSense, was there any opposition from within the company?
There was internal concern that using AdSense would decrease product sales on dinos.com. We explained that sales revenue had not changed since AdSense was implemented. In addition, we offered to stop using AdSense if we received any complaints from our users. In the end, sales levels remain constant and we did not receive any complaints. This shows that we’re providing useful information to our users and showing high quality ads

■ IA: Was it easier to implement AdSense than you’d expected?
It was very easy to introduce AdSense. It was very helpful that we could set up the ad categories that we wanted to block in the account interface. It was also good that we could block ads on a URL basis.

■ IA: What resources do you use to manage your account?
Since we don’t need many resources to manage AdSense, only one person is managing it.

We hope you enjoyed learning about another publisher’s AdSense journey. Stay tuned for more stories.

Posted by Kensuke Ishida, Strategic Partner Manager

Automotive brands were among the first advertisers to adopt Google+ and they’re making it pay off in big ways. Our research has shown that the path to buying a car is becoming more and more social, with consumers relying on information and advice fro…

Gmail: We are introducing a redesigned compose experience in Gmail that saves you time, uses less screen space, and is simple to use. The new compose experience opens a window at the bottom of your inbox. You can compose multiple messages at once, sear…

In March 2010, we released the Geocoding API V3 web service, adding features such as a correctness measure and recommended viewports, to give developers more tools for creating amazing mapping applications. At that time, we also announced our intention to shut down V2 of the API in March 2013.

We’re hearing from many developers that they need more time for the transition to the Geocoding API V3. So, we’re extending the deprecation timeline for V2 by six months. This means the Geocoding API V2 will continue to work until September 8, 2013 and we will shut the API off on that date. If you’re a developer and have questions about migrating existing applications to Geocoding API V3, please see our handy upgrade guide for help.

Today we’re also reducing the limit on Geocoding API V2 from 15,000 requests per day to 2,500 requests per day, which is equivalent to the daily limit on the Geocoding API V3. If you are affected by this limit reduction and require more geocoding quota, please contact our sales team.

If you have questions about these changes, Google and the broader Google Maps developer community can help. Please tag your question with the “google-geocoding-api” tag on Stack Overflow.

Maps for Business customers will be unaffected by this change and will continue to be subject to their purchased limits.

Posted by Ken Hoetmer, Product Manager, Google Maps API

Python 2.5 has a special place in the heart of any Google App Engine developer, as it was the first runtime we launched way back in 2008. Since then, both Python and App Engine have advanced a great deal.

A year ago we announced our support for Python 2.7, which brings syntactic and semantic improvements to the language and includes powerful features like threading and a large selection of third-party libraries.

Not only does Python 2.7 make developers’ lives easier, the runtime is extremely cost-effective. Our customers have taken advantage of features like concurrent requests to reduce their front-end instance costs by more than 70% while handling the same amount of traffic.

Not surprisingly, the Python 2.7 runtime has proven to be extremely popular. Just over a year after launch, more than 78% of active Python applications on App Engine are using the new runtime, and more are being added every minute.

As both Python and App Engine evolve, we must occasionally make hard choices about which legacy runtimes we should continue to support. Today we are announcing the deprecation of the Python 2.5 runtime. The deprecation period will follow the guidelines set in our terms of service.

What does this mean?

• We will continue to run Python 2.5 applications throughout the deprecation period. For most customers, upgrading to Python 2.7 is trivial as most elements of Python 2.5 are forwards-compatible with Python 2.7. We’ve prepared a handy migration guide that covers the steps to migrate in detail.
• If your application is still using the already deprecated Master/Slave Datastore, then you should first plan the migration to our more reliable High Replication Datastore, as the Master/Slave Datastore is not accessible from Python 2.7.
• Future versions of the App Engine Python Development SDK will display warnings to developers deploying updates to a deprecated runtime.
• Starting from January 2014, we will no longer allow new applications to be created using the Python 2.5 runtime.

We encourage all developers using Python 2.5 to consider migrating as soon as possible. We’re confident that the vast majority of our customers will find the upgrade straightforward and the benefits substantial.

If you’re considering migrating, here are some useful resources:

• Our summary of the new features we’ve introduced in Python 2.7 including additional standard libraries and third-party libraries that are supported.
• A handy migration guide, explaining the steps to move a Python 2.5 app to Python 2.7
• App Engine engineers give an in-depth discussion during Google I/O on the many improvements in Python 2.7, and how to quickly take advantage of them.

If you have any questions about this deprecation, we encourage you to contact us at .

-Posted by Andrew Jessup, Product Manager

Introduction: Earlier this week we launched the application for the Computer Science Summer Institute (CSSI), a three-week summer program for graduating high school seniors who are interested in studying computer science. To give you some more insight into CSSI, we’re bringing you stories from former CSSI participants through the My Summer @ Google series. Today you’ll meet Chyheim, a CSSI alum and freshman at the University of California, San Diego who is studying computer science. If you’d like to learn more about CSSI, visit our website and apply today!

Coming from a high school that did not offer any computer science curriculum, I saw CSSI as an opportunity for me to learn the essentials and to obtain the skills to succeed in college. The majority of CSSI was dedicated towards teaching us the fundamentals of computer science and programming. In order to get our feet wet we started by learning App Inventor, then we dived into HTML/CSS, Python, JavaScript, and eventually the combination of all of those with Google’s App Engine. On a typical day, we would break away from coding and participate in a few activities and workshops. These ranged from tips on creating a resume and on being successful in a technical interview to a Lego challenge and a social with Google interns.  One of the best parts of CSSI was the time I got to spend with my fellow CSSIers in the residence halls. On the weekends off from programming, we had trips planned for Saturdays and then Sundays off to ourselves.

Eventually it became time for us to put all that we had learned together and create a product. Essentially, everything the instructors taught was geared towards providing us with the ability to go on our own and build virtually any web application we liked. And that’s just what we did. Throughout the final week we worked to develop our very own web applications. My team created a web application called Gif.it. Using the Python Image Library and several tools in JavaScript, we constructed an application that allows people to upload pictures and generate their own animated Graphic Interchange Format (GIF) image.

Months after CSSI, I have come to realize that I owe a lot of my current success in college to the instructors and the mentors from CSSI. The exposure I had to concepts such as data types, conditional statements, loops, and arrays, has made it easier for me to grasp them in my programming classes. Aside from technical skills, CSSI has provided me with a large network of friends and professional acquaintances. I have made a number of friends that I continue to keep in contact with for support and advice. I have also gained mentorship from different Googlers that I met through CSSI. The bond that was formed between the students and the instructors made it incredibly hard to say goodbye once the program was over.

CSSI is definitely an opportunity you do not want to miss out on. Moreover, my advice for anyone who meets the qualifications is to APPLY. You should still apply even if you doubt your skill level. CSSI is all about giving students the opportunity to learn more about computer science, therefore only having minimum skills is perfectly acceptable. As long as you convey your passionate interest in computer science through the essays and the interview you should be in good shape. Don’t stress it. You have nothing to lose and so much to gain.

Interested in CSSI? Apply here!

Written by: Chyheim Jackson-Burgess, CSSI Participant

Posted by Dave Orr, Amar Subramanya, and Fernando Pereira, Google Research

When someone mentions Mercury, are they talking about the planet, the god, the car, the element, Freddie, or one of some 89 other possibilities? This problem is called disambiguation (a word that is itself ambiguous), and while it’s necessary for communication, and humans are amazingly good at it (when was the last time you confused a fruit with a giant tech company?), computers need help.

To provide that help, we are releasing the Wikilinks Corpus: 40 million total disambiguated mentions within over 10 million web pages — over 100 times bigger than the next largest corpus (about 100,000 documents, see the table below for mention and entity counts). The mentions are found by looking for links to Wikipedia pages where the anchor text of the link closely matches the title of the target Wikipedia page. If we think of each page on Wikipedia as an entity (an idea we’ve discussed before), then the anchor text can be thought of as a mention of the corresponding entity.

What might you do with this data? Well, we’ve already written one ACL paper on cross-document co-reference (and received lots of requests for the underlying data, which partly motivates this release). And really, we look forward to seeing what you are going to do with it! But here are a few ideas:

• Look into coreference — when different mentions mention the same entity — or entity resolution — matching a mention to the underlying entity
• Work on the bigger problem of cross-document coreference, which is how to find out if different web pages are talking about the same person or other entity
• Learn things about entities by aggregating information across all the documents they’re mentioned in
• Type tagging tries to assign types (they could be broad, like person, location, or specific, like amusement park ride) to entities. To the extent that the Wikipedia pages contain the type information you’re interested in, it would be easy to construct a training set that annotates the Wikilinks entities with types from Wikipedia.
• Work on any of the above, or more, on subsets of the data. With existing datasets, it wasn’t possible to work on just musicians or chefs or train stations, because the sample sizes would be too small. But with 10 million Web pages, you can find a decent sampling of almost anything.

Gory Details

How do you actually get the data? It’s right here: Google’s Wikilinks Corpus. Tools and data with extra context can be found on our partners’ page: UMass Wiki-links. Understanding the corpus, however, is a little bit involved.

For copyright reasons, we cannot distribute actual annotated web pages. Instead, we’re providing an index of URLs, and the tools to create the dataset, or whichever slice of it you care about, yourself. Specifically, we’re providing:

• The URLs of all the pages that contain labeled mentions, which are links to English Wikipedia
• The anchor text of the link (the mention string), the Wikipedia link target, and the byte offset of the link for every page in the set
• The byte offset of the 10 least frequent words on the page, to act as a signature to ensure that the underlying text hasn’t changed — think of this as a version, or fingerprint, of the page
• Software tools (on the UMass site) to: download the web pages; extract the mentions, with ways to recover if the byte offsets don’t match; select the text around the mentions as local context; and compute evaluation metrics over predicted entities.

The format looks like this:

URL http://1967mercurycougar.blogspot.com/2009_10_01_archive.html

MENTION Lincoln Continental Mark IV 40110 http://en.wikipedia.org/wiki/Lincoln_Continental_Mark_IV

MENTION 1975 MGB roadster 41481 http://en.wikipedia.org/wiki/MG_MGB

MENTION Buick Riviera 43316 http://en.wikipedia.org/wiki/Buick_Riviera

MENTION Oldsmobile Toronado 43397 http://en.wikipedia.org/wiki/Oldsmobile_Toronado

TOKEN seen 58190

TOKEN crush 63118

TOKEN owners 69290

TOKEN desk 59772

TOKEN relocate 70683

TOKEN promote 35016

TOKEN between 70846

TOKEN re 52821

TOKEN getting 68968

TOKEN felt 41508

We’d love to hear what you’re working on, and look forward to what you can do with 40 million mentions across over 10 million web pages!

Thanks to our collaborators at UMass Amherst: Sameer Singh and Andrew McCallum.

(Cross-posted from the DoubleClick Search blog)We at DoubleClick Search know that search marketing has expanded dramatically in scale and complexity over the years, and today, large search campaigns are practically impossible to manage using manual met…

With nearly 5,000 earthquakes a year, it’s important for people in Japan to have crisis preparedness and response information available at their fingertips. And from our own research, we know that when a disaster strikes, people turn to the Internet for more information about what is happening.

With this in mind, we’re launching Google Public Alerts today in Japan—the first international expansion of a service we debuted last year in the United States. Google Public Alerts is a platform designed to provide accurate and relevant emergency alerts when and where you’re searching for them online.

Relevant earthquake and tsunami warnings for Japan will now appear on Google Search, Google Maps and Google Now when you search online during a time of crisis. If a major earthquake alert is issued in Kanagawa Prefecture, for example, the alert information will appear on your desktop and mobile screens when you search for relevant information on Google Search and Google Maps.

If you click “詳細” (More info) right under the alert, you’ll see more details about the announcement, including the full description from the Japan Meteorological Agency, a link to their site, and other useful information like observed arrival times and wave heights for tsunamis.

And when you open Google Now on your Android device, recommended actions and information will be tailored to where you are. For example, if you happen to be in Tokyo at a time when a tsunami alert is issued, Google Now will show you a card containing information about the tsunami alert, as well as any available evacuation instructions:

We’re able to provide Public Alerts in Japan thanks to the Japan Meteorological Agency, whose publication of data enables Google and others to make critical and life-saving information more widely available.

We hope our technology, including Public Alerts, will help people better prepare for future crises and create more far-reaching support for crisis recovery. This is why in Japan, Google has newly partnered with 14 Japanese prefectures and cities, including seven from the Tōhoku region, to make their government data available online and more easily accessible to users, both during a time of crisis and after. The devastating Tōhoku Earthquake struck Japan only two years ago, and the region is still slowly recovering from the tragedy.

We look forward to expanding Google Public Alerts to more countries and working with more warning providers soon. We also encourage potential partners to read our FAQ and to consider putting data in an open format, such as the Common Alerting Protocol. To learn more about Public Alerts, visit our Public Alerts homepage.

Post content Posted by Yu Chen, Partner Technology Manager

A diverse workforce is critical in helping us build products that can help people change the world. That includes diversity of all life experiences, including gender.

Women were some of the first programmers and continue to make a major impact on the programming world today. We think it’s important to highlight the great work women are doing in computer science, to help provide role models for young women thinking about careers in computing.

Tomorrow is International Women’s Day, and as one of our contributions to the celebration, we’re proud to support Voices Global Conference, presented by Global Tech Women. As part of this 24-hour live streamed event, Google will provide more than a dozen hours of free talks featuring women working in computer science, beginning today. To access the full schedule and our ongoing broadcasts, see our section on the Voices website, which will be updated throughout the day.

The Voices Global Conference is the brainchild of Global Tech Women’s founder Deanna Kosaraju, who also started India’s Grace Hopper Celebration of Women in Computing in 2010 with grant support from Google. The India conferences, which provide a forum for women to share their professional and research work in computing, have grown rapidly, with more than 800 attendees in 2012. So when Deanna proposed this global, 24-hour streamed conference, we knew it was a great opportunity to help women and other audiences around the world learn more and get inspired about the contributions women are making to technology and computer science.

Our sessions will feature a range of material, from new episodes of the Women Techmakers series and interviews with women leaders like the head of Lexity India Mani Abrol, to discussions focusing on technologies like Google Compute Engine. For a sneak peek of the type of content we’ll be providing, check out Pavni’s story below, produced in conjunction with PBS’ MAKERS series. I’ve provided advice to many young people in India interested in studying computer science and pursuing their own dreams—so Pavni’s tenacity, coupled with the encouragement and support she received from her father, resonated with me. We’re excited to share her story and others like it alongside technical conversations and discussions on women in technology as part of this conference.

I hope you’ll join us for our sessions—and in the meantime, you can learn more about our efforts to support women at Google and beyond.

Posted by Beryl Nelson, Software Engineering Manager