Exploring Local
Mike Dobson of TeleMapics on Local Search and All Things Geospatial

Reference and Authority in Online Mapping

November 25th, 2010 by MDob

It occurred to me recently that many online mapping systems appear to be created by people who have read all the right books, but do not understand the concepts. (Bet you guys missed me, huh?) It appears that some of the online mapping websites have overlooked important aspects of accuracy, authority and reference. These items are of significant importance in mapping and would take a book to service adequately. Consequently, I am going to cut some corners while trying to describe my concerns in this area, realizing that it is a holiday weekend and many of you will have consumed enough tryptophan to cause a serotonin release resulting in the mother of all food comas.

As I have mentioned in previous blogs, maps are a generalization of reality and a measure of their goodness is how well they represent the reality they model. In a very simplified sense, the quality of geographic data used for mapping is some function of:
• completeness – presence and absence of features, their attributes and relationships
• logical consistency – degree of adherence to logical rules of data structure, attribution and relationships (data structure can be conceptual, logical or physical)
• positional accuracy – accuracy of the position of features
• temporal accuracy – accuracy of the temporal attributes and temporal relationships of features;
• thematic accuracy – accuracy of quantitative attributes and the correctness of non-quantitative attributes and of the classifications of features and their relationships.

(See ISO (ISO TC 204 for more information. This source might help those of you starting out )

At first blush, I would have to rate the world of online maps with less than passing grade on these measures. While I am willing to concede that many of the mapping sites might perform adequately on a measure of positional accuracy, we already have written about the industry’s inability to meet the test of reasonableness in respect to completeness, logical consistency, or temporal accuracy. Thematic accuracy, especially applied to geographic names and borders is something that we have not discussed in some time and given events in the news recently, it seems like a fine time to do so, unless you are Google Maps or is that Garbled Maps?

You may have noticed that I used the word “reference” above. The term has several meanings, but for our purposes we want to focus on the concept that a reference is a work frequently used as a source (The Free Dictionary). In turn, this suggests that we regard references as authoritative.

The notion of authority in the world of mapping is a complex issue, but at its base authoritative data are data produced by an entity authorized by a legal authority. For example, in the United States the authoritative source for geographic names used by Federal Agencies is the United States Board on Geographic Names, a body created in 1890 to maintain uniform geographic name usage throughout the Federal Government. While Google, MSN, or OSM can name a place in the United States anything they like, there is an authoritative name, or if there is not one, the BGN will authorize one. There is also a United Nations Group of Experts on Geographical Names (UNGEGN) that renders the UN’s opinion on the authoritative geographic names to be used by the countries belonging to the United Nations. Hmmm. Something seems wrong here. The UN isn’t the legal authority that establishes names for use within countries. What if the UNGEGN and a country disagree? How many authorities can there be? Plenty, it seems, or at least enough to confuse the issue.

Since there are so many “authoritative” sources for geographic names, perhaps we need to turn to the concept of trusted sources and trusted data. In short, a trusted source is one that publishes data gathered from authoritative sources. The trusted source often compiles data from a variety of authoritative sources and a provider is considered trusted when there is an official, documented process for compiling data from these sources and resolving differences when the sources conflict.

I suppose it is somewhat remedial to suggest that maps are designed to serve a purpose. For example, many online mapping sites serve as navigation aids, helping their users find locations and how to navigate to them from some other location. In addition, many of these same mapping services try to convey spatial information about a variety of themes in addition to data elements focused on roads, streets and addresses that may complement in the main theme of navigation. Curiously, several online sites focused on navigation also seem to be trying to establish their product as reference quality for uses other than navigation, such as use as an authoritative world atlas. In some ways, online mapping seems to be becoming the cartographic reflection of “Will it blend?” In other words, “Don’t try this at home!”

Unfortunately, most online map publishers are attempting to qualify as trusted sources, even though they, apparently, do not know this is what they are attempting to accomplish. However, since they do not publish their documented processes for compiling their sources and resolving conflicts, it is difficult to assess the validity of the processes and standards or their qualifications to be considered reference publishers. For instance, in respect to depicting country borders, online mapping organizations (or their data suppliers) have to gather data from authoritative sources or other trusted sources and attempt to make sense of the information they gathered. While anybody can make a decision, perhaps reference quality geographic data is an area where expertise is important.

In today’s internet-centric world, it appears that everyone else already seems to know everything there is to know about maps, but being slow afoot and something of an individualist in orientation, I have to work very hard to keep up with what is apparently common knowledge. Yep, I read a lot of stuff, trying to figure out what the online mappers are doing. Lately, for example, the internet news is full of stories focused on how Google is toiling mightily to keep its maps up-to- date, including keeping track of the positions of international borders. Even SearchEngineLand seems to be interested and may even be regarded as an authoritative news source for maps by some technophiles.

You know, life is good for Google, all they have to do when they screw up a map is to issue some marketing-speak about how hard it is to keep maps up-to-date, or that they have no idea where the source of the error came from and the world seems to scratch its collective head and wish them well. Can you imagine a spokesperson from National Geographic facing reporters and telling them “we have no idea where that border information came from.” Gosh, it’s just so difficult to get things right. Mapping is so difficult, but we try our (sob) best.” Nah, never going to happen. Only Google gets a pass when exhibiting incompetence in editing maps and using ignorance of map sources as their excuse.

So far this year, Google has: managed to irritate numerous governments with its “map country borders in erasable ink program”, nearly started a land war along a section of border between Nicaragua and Costa Rica (apparently Google thought this was OK since one of the countries did not have an army and could not mount an invasion in the case the border, as Googleized, was depicted incorrectly) and claimed an islet in the Mediterranean belonged to Spain, then belonged to Morocco, before finally calling it … “disputed”. How droll of them. In the process of trying to settle the issue, Google, according to the Associated Press, indicated that its goal was to be “neutral.”

Imagine that, being “neutral” in the world of mapping. Hmm. Last I looked, neutral was defined http://www.thefreedictionary.com/Neutral as, “not aligned with, supporting or favoring either side in a war, dispute or contest.” I guess, that means that Google is neuter when considered as an authoritative source. After all, the reason for these errors in borders is that Google did not do its homework. Well, actually that’s not right either, since all Google does is ask others who might be authoritative their opinion on the location of borders, as, it appears that Google does not have an authoritative opinion on map data. Maybe they just conflate all the recommendations and come out with an average? Well, it’s an approach a software engineer might take. Perhaps this is an opportunity for a new definition of authoritative, meaning, in this case, “We can say authoritatively that we don’t want to offend anyone.” Hmmm. Maybe Google should look at their depiction of the borders and the names they use to represent the intertwined geographies of Armenia and Azerbaijan?

By now, I suspect you think I am being unfair to Google, but would you accept Google’s own words on the subject? If so, take a look at this entry at the Google Latlong blogspot l. It starts out pretty nicely, indicating that the company has improved the quality of borders in Google Earth and Maps. The author continues:

“Making Google’s mapping tools as accurate as possible is a complex process, especially when a map’s accuracy has both quantitative and qualitative aspects. We receive spatial data of all kinds – imagery, boundaries, place names, etc. – from a variety of sources worldwide, and we review them carefully before integrating them into the best representation of a given location in Google Earth and Maps.”

Hmm. How do they decide the “best” representations? I’m not sure. I haven’t seen Google’s editorial standards. Have you? I suppose it does not matter, since I have saved the best for last.

In an another piece, the voice of Google said this, “Authoritative references: While no single authority has all the answers, when deciding how to depict sensitive place names and borders we use guidance from data providers that most accurately describe borders in treaties and other authoritative standards bodies like the United Nations, ISO and the FIPS. We look for the references that are the most universally recognized for each individual case. For example, in the case of “Myanmar (Burma)” ISO and FIPS each use a different name, so we include both to provide a more complete reference for our users.”

That’s somewhat reasonable.

Unfortunately, just a few sentences further along in the article is this gem “We work to localize the user experience while striving to keep all points of view easily discoverable in our products”. The last gem is followed by this one, “Carefully considering Google’s mission, guidance from authoritative references, local laws and local market expectations, we strive to provide tools that help our users explore and learn about their world, and to the extent allowed by local law, includes all points of view where there are conflicting claims.”

Hmm. Explain this to me again. Just what part of being authoritative depends on Google’s mission? And if the “guidance from authoritative references” does not fit with Google’s mission, what happens?

Next, if Google’s authoritative references do not agree with local laws then it appears that the local Google mapping site would elect to conform to the non-authoritative, but authoritarian local government’s point of view (although in this case local actually means “national”, say, as in China.) Good thing Google does not have a local mapping website serving North Korea. If it did, say goodbye to South Korea.

But how about that last part of the quote on “…local market expectations? Does that mean that if the advertisers in a local market don’t like the map, Google will change it and not represent the data the in the manner suggested by their authoritative sources? I guess if Google had any ad sales in Burma, the country would be called Myanmar not both Myanmar and Burma.

I suppose we can abandon the notion of Google Maps being authoritative, or, using our best market speak, we could note that “Google is flexible in its approach to mapping borders.” In non-market speak terms it appears that Google might actually consider utilizing a “whatever it takes” approach to satisfy local authorities so that Google advertising can continue to make the company gazillions of dollars. I’m not sure there is anything wrong with the approach, but it does produce a certain tension with the notion of authority and authoritativeness.

The problem with Google’s approach, in my opinion, is that the company appears not to have a set of unified standards for making its decisions and lacks a cartographic editorial standard based on the most authoritative sources of information. Based only on what Google has written about itself, which one assumes is authoritative, Google’s view of the best solution for representing a geographic issue is based on “what’s best for Google”. The real problem though, as a viewer of Google Maps and Earth, is how do you distinguish the representations that are authoritative from those that might be tainted by commercial interests?

Do you sense the tension between Google’s being a business whose mission is: “To organize the world’s information and make it universally accessible and useful” and being “authoritative”? Perhaps, Google might better be explained as an exhaustive source, but not an authoritative one. In fact, that is the path than many gazetteers take, loading up every name variant for a location, without showing a preference for one or another.

As curious as it may seem to you, the point of this blog was not to smear Google who makes adequate maps for most uses, but to highlight the difference between the law of big numbers and authoritativeness. Google for all its success and profitability may not understand that the role of maps for reference purposes is not always compatible with the company’s strategic plans for advertising or with its corporate mission to organize the world’s information and make it universally accessible.

I had a good laugh recently when one of my contacts was attempting to show me a screw-up in Google maps while using Bing Maps to prove his point. I asked, “What makes Bing authoritative?” Guess Steve Coast will have some additional responsibilities at Bing besides being chief architect. On the other hand, now that Bing Maps will be linked with that authoritative and poorly attributed tower of power known as OSM, why would they need authoritative sources? Just ask Fred, or Tim or Olaf (Jane might not be a good choice, since it appears that less than five percent of the contributors to OSM are female.) Or ask MapQuest. Of course, since MapQuest’s versions of OSM could have name and border changes every ten minutes this might not be a satisfying experience (not to mention you cannot know in advance if the data you are examining will ever stop changing as warring mappers practice the cartographic version of Whack-A-Mole.

And now for two loosely related items. I received an email from a settlement involving HP inkjet printers. Seems that HP was using colored ink and black ink to print black letters, using my ink more quickly than necessary. Other HP cartridges stopped printed on a date set by HP, even if the cartridge had ink remaining in it. Still other HP cartridges would send a message that they were low on ink when they were not. HP settled but did not admit guilt. Guess they are real stand –up kind of guys!

In another news item, a court case recently unsealed revealed that Dell was knowingly selling computers with motherboards equipped with bad resistors that would cause damage to the components of the computers. Dell’s solution was not to tell the customers but to let the boards and computers fail and then fix them. I had one of the problem computers, but did not know the problem. I went through four hard disks and two mother boards, along with hours of agony. All of this was fixed by Dell but when the machine went off its extended warranty I ordered a dell motherboard and now know that I inherited the same problem one more time. I had sworn off Dell’s, but in August I weakened and purchased a new Dell XPS 9100 with a Raid Array. It has now lost two disks, although Dell has replaced three disks just to be safe. Hmmmm. More stand-up kinda guys.

And now I find out that Google’s maps reflect local market conditions in the most recent version of “We didn’t want to, but they made us.”

I may have to rethink my stand on capitalism.

Well, Happy Thanksgiving anyway.

Some of the ideas on the concepts mentioned above in respect to authority and trusted sources were based on this excellent article “Authority and Authoritative Data: A Clarification of Terms and Concepts” by David Stage Fair & Equitable • February 2009 pp13-16 http://www.iaao.org/uploads/Stage.pdf

By the way, if you have not looked at the comments on my last blog on MapQuest, do so. Ant, a developer from MapQuest, answered some of my queries from the blog and his responses were both helpful and enlightening. Thanks, Ant.

Click for our contact Information

Bookmark and Share

Posted in Authority and mapping, Bing maps, Google, Google maps, map compilation, map updating, Mike Dobson, OSM

One Response

  1. Michael

    “But how about that last part of the quote on “…local market expectations? Does that mean that if the advertisers in a local market don’t like the map, Google will change it and not represent the data the in the manner suggested by their authoritative sources? I guess if Google had any ad sales in Burma, the country would be called Myanmar not both Myanmar and Burma.”

    Is there any evidence Google is editing the core map content based on the desires of local advertisers, or is that speculation?

    To me “local market expectations” could simply mean that if people in one country call something “The Persian Gulf” and people in another call it “The Arabian Gulf” then they show the local name. The way I see it Google is just making a map I can use to find my way to my buddy’s house, they aren’t trying to get involved with the Kashmir conflict or anything.

    Michael – I am sorry but somehow I lost track of this comment. My apologies for the delay in posting it.

    In respect to your comments, the representative from Google is the one who indicated that the company would be sensitive to local market expectations. Note that it was not stated as sensitive to “local user expectations.” I think your views on maps misses the point that map editors (just like us bloggers) have an editorial perspective. However, when the perspective is influenced by sales and market factors, the map may not reflect geographical or geopolitical realities.

    Google is attempting to become a reference authority in mapping and as such, its editorial responsibility falls far beyond making sure the street names help you to find your way to a buddy’s house. I think you missed the message of the blog and suggest you read it again.