GEOG 583
Geospatial System Analysis and Design

Technology Trends: Volunteered Geographic Information


An increasingly common trend in GIScience is the development of data sources that are provided by communities of volunteers, rather than for-profit businesses or government entities. Volunteered Geographic Information (VGI) is spatial data that has been contributed for free by volunteers.

A wide range of VGI efforts have taken shape over recent years, but the one that has received the most attention so far is OpenStreetMap (OSM), which is a project intended to develop a free base map for the world. By relying on VGI contributions, OSM is able to develop a map product that competes nicely with commercial maps while having none of the restrictions associated with for-profit mapping enterprises.

It's important to consider that there are multiple types of VGI - for example, one form of VGI would be adding road segments to Open Street Map by using a GPS and uploading the data to the OSM site. Another type of VGI would be your willingness to share your location information on Twitter so that users of the Twitter API can find your Tweets when doing spatial queries. In the first case, the volunteer is taking an active role in data creation. In the second case, the volunteer is not really doing anything beyond normal behavior and is simply allowing someone else to view and use automatically recorded spatial information.

The first short video I'd like you to review provides an overview of how Tomnod (recently acquired by DigitalGlobe) integrates principles of VGI into image analysis.

Please watch the video Tomnod joins DigitalGlobe (2:55).

Click for Transcript of Tomnod Joins Digital Globe Video

Hi, my name is Shay Har-Noy, former CEO of Tomnod and now I’m part of Digital Globe.

Why did you decide to start Tomnod?

Tomnod was born out of commercial imagery industry. From our beginning, searching for the tomb of Genghis Khan as part of UC San Diego, we got a grant from the GOI Foundation, where they gave us imagery - satellite imagery of remote northern Mongolia that we used in our search.

Since then we've been establishing stronger and stronger relationships with the players in the commercial imagery market. So, for example, we worked with Digital Globe last year when the uprising was going on in Syria, where we were able to leverage Digital Globe imagery as it was being collected and apply our crowd-sourcing algorithms, our crowd-sourcing methodology, in order to extract value, extract information from that imagery.

What we're really trying to do is get close to the industry, get close to the players with the satellites in order to help them unlock the power that’s in the imagery.

How do you apply crowdsourcing to imagery?

The most promising applications of crowdsourcing combined with satellite data are the ones that have the most pressing time constraints.

For example, last year we had the unfortunate event of two missing climbers in Peru. We teamed up with Digital Globe and with GOI in order to try to get the most recent satellite data of that region. And over the course of the next four hours, we had 800 people comb through that entire mountain range, the entire region, searching for small disturbances in the snow. Seven hours after the imagery was captured, we had our final clues. The crowd had identified two subtle tracks in the snow, two subtle tracks leading to the edge of a cliff.

And so this story does not have a happy ending, but it really highlights the power of the technology, the power of the imagery, and the power of the user.

What are you looking forward to most working with Digital Globe?

When we started Tomnod we had a vision of making satellite imagery accessible and its analysis scalable on a larger scale. We wanted eyeballs on every pixel that's collected from space. That's a tall order. That's hard to do. When we joined Digital Globe we took a huge step in that direction. No longer are we just obsessed with resolution and accuracy. Right now we're looking into accessibility. And that requires two things. That requires distribution and scalable analysis. We want to empower the end user. I'm looking forward to a day when anyone can access satellite imagery, anybody can extract information from it in order to make better decisions.

My name is Shay Har-Noy, former CEO at Tomnod, currently director of research and development at Digital Globe. I’m proud to be here. Thanks for listening.

The second video I'd like you to watch focuses on a critique of VGI specifically focused on how sustainable VGI efforts will be in the long term. In essence, is it reasonable to count on VGI efforts to develop high-quality data sources in the future, or is it a short-term fad that will fade away?

Please watch the video Is volunteered geographic information sustainable? - Steven Feldman (19:37).

Click for Transcript of Is Volunteered Geographic Information Sustainable? - Steven Feldman

PRESENTER: Fantastic. So-- Steven Feldman.

STEVEN FELDMAN: OK. I spent a lot of time in saying--


STEVEN FELDMAN: Can you hear me now?


STEVEN FELDMAN: OK. Well first of all, let me say thank you to OpenStreetMap for inviting me to speak. I'm a bit of a paleogeographer, but I spend a lot of time selling GI solutions to local government. And recently, the last couple of years before I got out of doing that, people would say, talking about OpenStreetMap in local government-- you heard James, yesterday. And the question they asked was, is it sustainable?

And they didn't really know quite what they meant by sustainable, but they really were just worried that it was a flash in the pan. So what I wanted to do was talk today of whether volunteered geographic information is sustainable, what that might mean, what motivates the OpenStreetMap community, perhaps-- because that might be a factor-- and where stuff might be going.

So I'm going to just look for a while at sustainability and what it might mean. I'm going to look at what's going on in terms of participation. And Steve-- you're looking at me-- used half my slides already, so I'll whisk through them. I'll talk about some stuff on motivation, and then I'll do a little bit of future case. So help me out, here. Let me just get an idea of who's in front of me. How many of you contribute to OSM frequently? Right, that's a hard core audience, isn't it? How many of you are moderate? Not many? And how many only once or twice? OK. Now, how many of you earn your living from something to do with geographic information? And how many of you are you earn your living from some other kind of IT? Right.

So what we've got here-- Oh, and just answer me one more question.


OK. So let me see if I can play this back to you. Remind me if I'm wrong. We've got a hard core audience of OpenStreet members. You're mainly in the frequent category. One heck of a lot of you earn your living from either IT or GI, and not that many of you are married. OK. Fair enough. That's given me an idea. We'll see what that means later on, hey. Oh, anyone got kids?


STEVEN FELDMAN: Right. I didn't correlate them, the married and the kids, and I'm not going there.


So OK, what do you want from a street map? Right, I don't know. Let's see. I mean, [INAUDIBLE] has covered this to some extent. There's some sense of completeness. We don't really need to know what complete is. We all know we want some sense of completeness. Whoa. We want some level of accuracy. Again, you know, let's not go into it. [INAUDIBLE] has has done it really thoroughly. We certainly want currency. I think currency is a really important point, and I'll come back to it in a while.

And put that lot together, and you've got some sense of we want something that's reliable, right? Reliable-- we know what it is, and we know that it's going to be like that in the future. So that thing would be a near perfect map. This near-perfect map would be complete, it would be accurate, it would be up-to-date all the time, and we would have confidence that it would keep being right.

So how does a street map become sustainable? Like, what are the things we need to do to make it sustainable? Well, it's down to two things. It's down to people, and it's down to processes. And it's not for me to talk to you about process, today. You heard it from GEOfabric yesterday, you heard Steve talking about it. I'm going to talk about people, lots of people, right?

So massive numbers of contributors. We hear 130,000 contributors. Steve's talking about 1 million contributors next year or the year after, also know as the contributors. But does that mean we're soaring to magnificent new heights, or are we just about to go under the tipping point and skittering down the hill? Is there contributor fatigue? Right, yeah. Are people just doing it for a while, and then dropping off? You know, what Steve was saying would suggest that there is some fatigue going on.

So you might know this. This is Gartner's famous hype curve. And please, I'm not suggesting that OpenStreetMap is hype, but it's an interesting curve. You get up to this peak of inflated expectations. Then you drop down into the trough of disillusionment, and then you get back up, again, into this nice, steady plateau, right.

Now, that could be a pattern for OSM contributors, or we could have something like that, which would not be a good thing, except, if you're going from 100,000 to 1 million people, and you keep moving these peaks along for each of those people, you actually can get a nice flat, or even growing, line, even with the people dropping off. But it's probably not the most secure way of building, necessarily, the map.

So I did some work. I looked at some of this stuff, and I discovered, in North London, where I live, that actually you've got a tiny number of people. And there's Steve Chilton, sitting over there, who's one of the top three mappers in North London, right? And there you've got the famous long tail, which we've seen, right? 5% of the contributors in North London with 81% of the edits.

But what's interesting is when you look at this, which is graphing it over time, you see all of these bumps and different colored lines and different people peaking at different points. That suggests that even though you've got a small number of people, they're coming in at different times. And when you look at it overall, what you're seeing is quite a sustainable level of edits going on in North London.

Now, I thought I'd go and look at Redlands, California. I'm sure you all know what is based there. I thought there'd be a few map geeks, there. So let's see what's going on in that space. Right, well you see is it's really down to two people, maybe three, right? There's the graph, right? One big peak, and there you've got maybe a trough of disillusionment going on at the moment.

So then I wanted to look at Belfast. Two different communities, probably some tensions about where people go riding around on their bicycles with GPS-- maybe, I don't know. Maybe it's all down to one guy-- same kind of thing-- 84% of contributions, 4 people. A bit peaky, there. There is a second peak there, which might be encouraging. These are all 3-year graphs that I pulled there.

And finally, I thought, I'd better look at Amsterdam, because that's where we are. So forget the contribution number indeed, because it's going to skew everything if you put that in, and look at the users beyond that, and what find is that 76% of the edits are from 5 people. You've got a pattern, here. I won't go on. So here are a couple of observations. Most cities are dependent on a small number of contributors, but there is something evidence of user fatigue. And I think it's something that you do need to look at, perhaps, is that large cities are not more sustainable in terms of recruiting new contributors than small cities. So does this matter? What the heck.

Well, I think it does matter. Updates are a bit like looking for a needle in a haystack. You know, it's dead easy when you've got all the fun of going out there and mapping a whole new area. There are great unmapped parts of the universe that you can go and map. But when you've got go back and find those updates, as Steve was saying before, that's more challenging. It's less fun, and that is going to test the motivation and commitment, perhaps, of the foundation members, the community members, as well.

So [INAUDIBLE] happily pointed me to this article, which was more motivates Wikipedians. It's quite a good article, because it looks at classic factors that motivate people and then tries to correlate them with actual levels of contributions. So there's two separate things going on. There's motivation-- what turns you on. And there's what makes you actually contribute.

And what I'd like you to do, as I quickly run through these, just think about your own motivation for contributing to OpenStreetMap. You know, you're a high intensity group of people in terms of your contribution. So when I go through these things, what is it, which one's have been [INAUDIBLE] So is it values? Right, there's something about doing this that [? jives ?] with your values. Is it the social element? Is it understandable learning? Right? We all learn something, perhaps, from doing this.

Is it that it some way impacts on your career? You know, is it for job motivation? Is it some sort of emotional self-protection, that by doing this, you're compensating for something else that's going on in your life, right? I'm not making these up. This is serious stuff. Is it-- oh, it couldn't possibly be that value, there, right? It can't be that. I'm sure it's not.

There are also two other factors that are less classic for volunteering, but do apply to projects like this. Is it about fun? Or is it about ideology? And when I'm talking about ideology, it's not just politics. It could be a sort of really strong commitment to open basic principles, or something like that. So you've got eight factors, there.

OK. It's time for you to do a little bit. You've been sitting on your backsides for ages. We're going to do a little bit of exercise, now. Steve, can you help me, here?


STEVEN FELDMAN: All right. Because what I want you to do is I want to get people to just indicate which ones. And I need a second person just to give me a sense of which are the top three values. OK?



AUDIENCE: [? But on me. ?]

STEVEN FELDMAN: You're [INAUDIBLE]. That's fine. So am I.


OK. So if you think that values are one of the-- you can have more than one, so if it's a big contributor, stick your hand up for values. OK. That's not very much, is it? Whoa. Go back. If you think that it's social factors, stick your hand up. About the same as values. OK. If you think it's the learning experience, hands up. Wow. That's a little bit higher, I guess. Do you agree with that?


STEVEN FELDMAN: If it's protecting something else that's going on in your life, the compensation, stick your hand up. I didn't figure--


OK. We'll talk later. We'll talk later. Right. So if you think it's about enhancement, boosting your ego, and stuff like that, stick your hands up, sure. Wow.


OK. Now, is it about fun? OK. I thought so. And is it about ideology? OK. Right. Now, that's really interesting, because now-- so I'm going to say, and Steve, tell me if you agree, that we have values, and we had-- we had values, social, and understanding about the same. We were low on protective and enhancement. We were very, very high on fun, and very high on ideology. Brilliant. OK. So here are the top four motivations that Oded Nov found in his study of about 370 highly active Wikipedians, a similar sort of cross-section to this one.

And they were, guess-- look at that-- fun, ideology, values, and understanding. Now, learning-- correct. A couple of interesting things, here, is he said that ideology, actually, when they did the test and looked at the correlation of ideology and contribution, ideology does not correlate with contribution. You all are strong on that factor, but it's not apparently correlating with the level of your contribution, but fun, values, and understanding are.

I think there's another factor here which makes OSM very different from Wikipedia is, when you go out and map, you tend not to do that as a solo activity. A lot of mapping is done as a social activity. And I think that you gave a much higher rating to social than the people who edit Wikipedia in their study of [INAUDIBLE] So those are the factors. So what's all this mean for the OpenStreetMap community?

Well, if the map's going to be sustainable, what I think you have to do is I think you have to look at what are these factors that motivate you to contribute and work out how you can build on those factors to engage that large, long tail that currently aren't contributing. It would seem to make sense. You know what turns you on. How are you now going to make that work for a lot of other people.

So you've got to build a community. We all know that. You're going to do that by having fun. I think you're doing this already. It sounds great fun. Note, the great pictures that Steve had of kids going out and mapping. I'm going to see if I can persuade my 12-year-old, who is really keen on his bike, at the moment, to stick a GPS and go out and do a little bit, and see whether I can get him involved.

And value is really important. As you remember, you've got to make a difference. And I think that, for me, the thing that most inspires me about what you do in OpenStreetMap is when you go out and map the places that nobody else wants to map, right? That's the bit that's really cool. Right?

Doing the same thing as [INAUDIBLE] is interesting, and I understand the politics of it. Mapping the places that nobody else wants to map is really great. So, and is learned, right? Just a last thought for you-- is it time to learn something from some of the people who have [INAUDIBLE] found open source foundations. Is it time that we ought to actually be maybe inviting the big guy to come in, and he is big.


Now, I don't know who that big guy is. Could be-- I mean, one possibility is it's the big G. Another possibility is that you invite a friend. These guys have been your friend for a while, maybe you ought to be talking to them. I don't know. These are just observations, now. I'm rambling. Maybe you need incentives, OK. That incentive could be about money. It could be people actually getting some kind of reward for doing something. It could be non-monetary stuff. I mean, people want recognition, and people want approval.


So in conclusion, OpenStreetMap, you're doing a great job. You've got a great community. I think that you are coming to a crossroads. I think there are some things that you're going to decide on. I'm sure you're going to make some good decisions. So back to my initial question-- is volunteered geographic information sustainable? Well, if you keep doing these things right, keep doing the things that you're doing, the answer surely has to be yes.

And oh, that's interesting that that's up. That's me. Thank you to ITO for those great maps. Thank you to flickr for thousands of great pictures. That's me, again.


PRESENTER: Whilst Gary Gale is getting set up, we can take a couple of questions. So who's got the first question?

AUDIENCE: I think while I disagree with your analysis about how the number of users contributing is peaking like that, I think it's kind of inevitable that you're going to get a relatively small number of people fill in the initial map for an area, it's going to involve lots of edits painted on a blank canvas.

And then I think it, also, that, [INAUDIBLE] you're going to get a somewhat larger number of people, making a small number of edits, where they make corrections, editions, as time goes on. So I don't think that's a bad thing. I think it's kind of inevitable, in the way that the map starting with a blank canvas works.

Also, in terms of motivations, I think I might not have it at all, because I've got two motivations which don't apply to Wikipedia, that wouldn't have asked about Wikipedia, one being profit. The reason I started out doing mapping is because I wanted to publish at the end of it. And it was a convenient tool to be able to do some of the work that I couldn't have made just on my own. But having the rest of the tool set there to help me was good.

And secondly, exercise, which you don't get, with a Wikipedia user, and that working from home, you've got the [INAUDIBLE] of the house with a [INAUDIBLE] target. So I suspect there's other motivations, as well. That's a question about--

PRESENTER: OK. That's one question. Steven, what do you say to that?

STEVEN FELDMAN: Look, that's a great statement. I mean, I think you make a really good point about the different profile or the later edits in the field. And I think it would be really good if you could do a little bit more analysis, and look at what those small number of edits, all those people who make 5 and 10 edits, what are they actually doing?

I mean, from the IT ITO stuff that I had, I couldn't work that one out, but maybe, that's an interesting-- you could. But I'm certainly not saying it's not sustainable. I'm just showing you some patterns and playing with a hype curve and trying to interest you for 20 minutes.

AUDIENCE: I haven't got any numbers to support it, but I do watch what's going on in my area, so I do know what, sort of informally, what sort of the small edits are doing. and they are making corrections, they are adding detail. And so they're using the fact that the structure of the map that's already there to improve what's there.

STEVEN FELDMAN: In that case, it's going to be sustainable.


PRESENTER: Good points around motivations. Gary, were you ready to roll?

Finally, I'd like you to take a quick look at an example where VGI worked wonders in a real world situation. Shortly following the Haiti earthquake in 2010, volunteers worked rapidly to compile and develop high quality road network data using the OpenStreetMap framework. A short blog article summarizes this effort, and its impact is punctuated by this very short but dramatic video (note that there is no narration).

Please watch the video OpenStreetMap - Project Haiti from ItoWorld (:31).

OpenStreetMap - Project Haiti from ItoWorld on Vimeo

Joining the Discussion

Post a comment in the Lesson 3 Technology Trends Discussion in Canvas.