Links: Zillow News, Virtual Volcanology, Design in GE, GearthHacks News, Ski Time, VE Geocoding

  • Zillow News – Popular real estate searching site Zillow mentioned on their blog yesterday the release of their database of shape files for boundary lines for 7000+ neighborhoods (apparently the neighborhoods – not the individual properties) around the US. The release is made in the form of SHP files and they are released under a Creative Commons license. According to some blogs, the license allows this: You are free to use the files in this database in applications as long as you attribute Zillow when you use it. You may also make your own changes to the database files and distribute them, as long as you provide them under the same kind of license and give Zillow attribution. I’m hoping this means someone will take the shape files and make them available as a dynamic network link in Google Earth. You can find the Zillow data here.

  • Virtual Volcanology – John Bailey of the Alaska Volcano Observatory (and one of the organizers of the AGU virtual earth sessions), was invited and gave a Google Tech Talk last month. His talk was titled: Virtual Volcanology – Exploring volcanoes in Google Earth. You can watch it now in YouTube. I particularly liked the short video he presented which starts about 42:00 into the talk.

  • Design in GE – Richard Treves of the University of Southampton in the UK has posted the presentation he gave at the AGU conference last month. His presentation makes a compelling case to put some thought into the basic design of your visualizations.

  • GearthHacks NewsGearthhacks.com has undergone some design changes. Mickey contacted me to let me know he has created a new tool called Yourmap to help churches create their own basic maps and embed their map on their web sites. It’s a mapping mashup of Google Maps like many others,but the difference is that it is geared toward simplicity and it also supports Google Earth output prominently.

  • Ski Time – A skiing enthusiast blogs about useful tools in Google Earth for skiing and encourages folks to remember to take their GPS showing the example of GPS tracks I did for Google Earth.

  • VE Geocoding – Microsoft has blogged about an article in a Florida newspaper which did some tests of Virtual Earth’s geocoding capability compared to Google Maps. According to the newspaper in their tests Microsoft trounced Google by showing the proper location for addresses in 341 out of 400 cases. The newspaper says this clearly shows the superiority of Virtual Earth over Google Maps. However, as one reader pointed out in the comments: “Regarding geocoding accuracy: It’s not fair to do this in one small area. Both GE and VE often get their data from local vendors, and for this area Microsoft probably just happened to get the data from the best vendor, whereas as far as I can tell, GE just uses interpolated addresses along the street centerlines for this particular area. As with their aerial imagery in some areas one is better than the other, and vice-versa.” The Microsoft PR machine is in action. Here’s some other interesting posts in recent days touting Virtual Earth uses in areas mostly dominated by Google Earth or Maps already: Local news stations, MyWeather site, YellowPages.

About Frank Taylor



Comments

  1. Sorry, but I’ve got to take issue with your comments in regards to VE Geocoding.
    This is old news. In April of last year MS announced they were using ‘rooftop’ geocoding as apposed to street level approximation for addresses in US & Japan. MS geocodes better in those countries than Google. In time I’m sure Google will update their geocoding to match, but as it stands today VE will usually give you a much more precise location that Google Maps does. I would say it’s not the Microsoft PR machine working overtime here, but rather it appears to me like the Google PR machine is in full spin mode.

  2. Just a quick update on YourMap — I’ve been working today to build geocoding into the script, in an ongoing effort to make it as easy to use as possible. I’ve got that going now, so hopefully that is helpful for some people.
    Also, Frank — Thanks for the providing the link to John Bailey’s presentation. I haven’t finished watching it yet, but it’s quite good. I wouldn’t have known about it without that link, so it’s much appreciated.

  3. Ye ask and ye shall receive. Zillow Neighborhood boundaries for Google Earth.
    Blog post: http://www.gelib.com/zillow-neighborhoods.htm
    Direct link: http://www.gelib.com/maps/ZillowHoods/ZillowHoods_nl.kml
    Cheers…

  4. John Franham says:

    Typo alert
    South Hampton is one word Southhampton .

  5. Frank, I could not agree more about the MS PR machine.

  6. Not to mention, anyone who would release a tool like this — and not provide geospatial support for the formats ingested — is troubling, if they think they’re going to inaccurately PR their accuracy.
    http://dev.live.com/virtualearth/mapcruncher/
    I had actually asked Research if they planned on including support for a geospatial format, because I thought it would be a thousand times more useful if that were the case. The response I received was a rather profound ‘duuuuh.’

  7. Thanks for linking to my post on geocoding accuracy in West Palm Beach.
    Minor correction: The post is on my personal blog, which is not affiliated with my employer. I happen to work at a newspaper, but this was not an article written for or published by that newspaper.
    Clarification: I did not actually write that our test, as you put it, “clearly shows the superiority of Virtual Earth over Google Maps.” What I did write was that, when it comes to geocoding accuracy, “Google isn’t very good at it, at least compared to Virtual Earth, and at least in Palm Beach County.” I don’t see how one can dispute that, given our results in Palm Beach County.
    As I responded to the commenter who said it was not fair to compare geocoding results and imagery recentness in one small area, that one small area is the only part of the globe I happen to care about for development purposes. The title of the post, after all, was “Google Maps vs. Virtual Earth: A geocoding accuracy showdown in West Palm Beach.” That “West Palm Beach” part is all that matters to me. Google might blow VE away in Dallas and Helsinki, but that doesn’t do me any good in South Florida.
    Bottom line is that we use the best product for the job. We’re using Google Maps AND VE in our current project, for example, because they excel in different areas. And we’re doing all this work on Mac notebooks, a Linux development box, and a Windows workstation for GIS. No PR machines at work here.

  8. William,
    What I think you should point out in fairness to both parties then, instead of labelling the application itself as ‘inaccurate’ — is that both applications are in fact geodetically accurate in their technologies and how they handle geospatial data. It is the data itself that you’re placing in question, and it may be the case that depending on the data accuracy for a given location, or the accuracy of the data that was available to them at the time of acquisition — will or may have subtle discrepencies between the applications as a result of whatever data inaccuracies occur.
    That then, is distinctly relative to a data-provider issue — not an application issue.
    Keep in mind that this is one of those ‘things’ that keeps popping up which has led to some brow-raising on the part of those of us who are professionals in the industry. And unfortunately, that type of inaccurate or less-than-informed criticism has also seemed to historically remain consistant from a direction that’s become rather predictable in nature.
    That is my final opinion on the matter.

  9. @Daniel: I can understand your geospatial professional’s perspective. But from my user and development perspective, data provider and mapping application in this case are one in the same, as they can’t be separated on my end.
    All I care about is putting my pushpins on a map as close to their actual location as possible. If the best way to do that is dispatching a kid on roller skates in response to every single geocoding request, then sign me up for an enterprise license, regardless of whether he’s selling his data to Google, Microsoft, or the Foot-Powered Geocoding and Typewriter Repair Company of Kenosha, Wisconsin.
    Google has one black box, and Microsoft has another. We fed our addresses into those two black boxes and compared the results to the actual, real-world locations we already knew. At this moment, in our little corner of the globe, Microsoft returned better results in 341 of 400 cases. The degree to which I admire or loathe one compare or another isn’t going to change those results.

  10. @Daniel: You are 100% correct when you point out the inaccuracies are the result of the data rather than the application. “Garbage in, garbage out” as the old mantra says. What we (as geospatial professionals) need to understand is that the vast majority of people don’t care. I can’t count the number of people who tell me Google Earth is more ‘precise’ that Virtual Earth because Google has higher resolution imagery at the location they are looking at. Trying to convince someone otherwise is a lost cause. Have you ever attempted to explain to someone why Google Earth (and I assume Virtual Earth) aren’t as precise as most people think because they use a spheroid to model the earth rather than a geoid? No one cares. As long as what they are looking for lines up with the image underneath they think it’s great.
    We need to look at things from a different perspective. For example; Do you really care why one search engine returns better results than another? I know I don’t. I just use the one that finds what I’m looking for.

Leave a Reply