Archive for April, 2006

We re all for gender equality, so ever since we examined the lovely legs of Search, we ve been hankering for a taste of the hottest torsos, too. When buzz on aaron carter shirtless and

We have identified and corrected the ReportService bug we reported yesterday. The ReportService has been functioning normally for the last several hours. We apologize again for the inconvenience and are working hard to ensure that this type of situat…

Show your visitors where you stand with the Google Maps API. The Maps API allows you to embed geographical information into your site using JavaScript. You can add overlays to your map including markers, polylines, and shadowed information windows. It …

We ve received inquiries regarding the user agent string that the Windows RSS Platform uses when making HTTP requests to servers. This is particularly interesting to services that provide statistics to their users indicating how many users are reading content via web browsers vs. aggregators. There are two distinct instances where a user agent string is used: The user is not subscribed to the feed. The user navigates to a feed and IE7 presents a preview of the content. The user is subscribed to the feed. The RSS Platform retrieves the feed content on a schedule or on demand . In the first case it s the web browser that is making the HTTP request and hence the IE7 user agent string is used. For example, on Windows XP, this string looks like this: nbsp; Mozilla 4.0 compatible; MSIE 7.0; Windows NT 5.1 In the second case it is the RSS Platform that makes the nbsp;HTTP request and hence the following string is used on Windows XP: Windows RSS Platform 1.0 MSIE 7.0; Windows NT 5.1 and on Windows Vista: Windows RSS Platform 1.0 MSIE 7.0; Windows NT 6.0 The RSS Platform user agent string reflects on the version of Windows on which it is running, and the version of IE that is present on the system. Note 1: The current beta 2 build of nbsp;the RSS Platform uses nbsp;a slightly different string, nbsp;that is not compliant with the HTTP spec. The difference is that spaces are used in the token instead of dashes. nbsp;Future builds will use exactly the string shown above. Walter vonKoch

The 33rd Annual Daytime Emmy Awards are happening tonight, and will boast a historic first.This is the first year that an award will be presented for original entertainment programming created specifically for non traditional viewing platforms, including computers, mobile phones, iPods,…

By Denise Gamboa, Associate Product Marketing ManagerA unique feature of Google Base is the ability to create custom attributes and item types. This gives you an opportunity to include relevant information, no matter how obscure, detailed or specialized it is, about the content you want to share. For those of you who are already submitting data through Google Base, you ve probably noticed that we have a finite set of item types to choose from and that for each item type, we have a finite set of suggested attributes. Though you always have the ability to create your own, we d still like to present relevant suggestions for each item type that we call out.Because you re the experts in your fields, we d like you to tell us about what item types you d like to see in the Choose an existing item type: drop down on the post an item page. We d also love to know what attributes you think are most important for each item type. For example, if you re a photographer with interesting photos to share, license or sell, let us know what attributes make the most sense for your industry. Should we include print size, licensing information, place, camera etc? Or, if you re a collector of African art, what details do you and other collectors think are important when you evaluate your wares? Let us know by posting to the Base forum or via email at base feedback google.com. We d love to hear from you.

Posted by Franz Och, Research ScientistBecause we want to provide everyone with access to all the world s information, including information written in every language, one of the exciting projects at Google Research is machine translation. Most state of the art commercial machine translation systems in use today have been developed using a rules based approach and require a lot of work by linguists to define vocabularies and grammars.Several research systems, including ours, take a different approach: we feed the computer with billions of words of text, both monolingual text in the target language, and aligned text consisting of examples of human translations between the languages. We then apply statistical learning techniques to build a translation model. We have achieved very good results in research evaluations.Now you can see the results for yourself. We recently launched an online version of our system for Arabic English and English Arabic. Try it out! Arabic is a very challenging language to translate to and from: it requires long distance reordering of words and has a very rich morphology. Our system works better for some types of text e.g. news than for others e.g. novels and you probably should not try to translate poetry … but do stay tuned for more exciting developments.Update: We ve just opened a discussion forum for all topics related to machine translation.

Posted by Franz Och, Research ScientistBecause we want to provide everyone with access to all the world s information, including information written in every language, one of the exciting projects at Google Research is machine translation. Most state of the art commercial machine translation systems in use today have been developed using a rules based approach and require a lot of work by linguists to define vocabularies and grammars.Several research systems, including ours, take a different approach: we feed the computer with billions of words of text, both monolingual text in the target language, and aligned text consisting of examples of human translations between the languages. We then apply statistical learning techniques to build a translation model. We have achieved very good results in research evaluations.Now you can see the results for yourself. We recently launched an online version of our system for Arabic English and English Arabic. Try it out! Arabic is a very challenging language to translate to and from: it requires long distance reordering of words and has a very rich morphology. Our system works better for some types of text e.g. news than for others e.g. novels and you probably should not try to translate poetry … but do stay tuned for more exciting developments.Update: We ve just opened a discussion forum for all topics related to machine translation.

Posted by Jon McAlister, Software Engineer I ve always been a big fan of both Google News and Google Suggest. So in my 20 time, I ve worked on a way to bring them together, and I m now happy to report the launch of Google Suggest on Google News, wh…

Here ye, here ye… Bloggers publishers, Bloglines can now accept pings directly from you in two ways. Read more at http: www.bloglines.com services api. If you already ping Ping O Matic, or if your blog is hosted with LiveJournal, TypePad or MovableType, no need to lift a finger. We ve already got you covered. Bloglines readers, you too will benefit. Blog posts from publishers who choose to ping us will show up within minutes, sometimes seconds, of being posted. Your Bloglines Team