Minimal requirements - know what you want the information for...
Yann,
I really appreciate that you raised this very important point right at the outset, "Some minimal requirements for geo-mapping the end user should know what he/she would like to see on the map." We can't get to where we're going if we don't know where we want to go. In human rights work - that requires us to have a VISION. Can we even imagine the way we want the world to be? And if we can, what information do we need to help us get there?
You also made a great point about the need to have a system - and the idea of developing a common data set for human rights violations. In January 2010, New Tactics will host a dialogue featuring databases for human rights advocacy and work. I would like to see how that discussion can link back to geo-mapping and your idea of developing a common data set.
There are a couple of longstanding human rights databases:
HURIDOCS has some great tools and examples in their training logbook that might be of interest to you and others.
Martus that provides an encrypted information management system to protect sensitive data.
I'm very interested in seeing how such human rights data systems can be utilized to render geo-maps (taking into consideration the critical points raised already in this dialogue, especially ensuring that people are not further endangered or their families/communities compromised.)
Nancy Pearson, New Tactics in Human Rights Program Manager
OpenEvsys is a human rights case management system, based on the Event Standard Format. The system was built for HURIDOCS as per their requirements, by Respere. OpenEvsys is a Free Open Source project licenced under AGPL v3 - we are currently in the process of building the community around this. More detailed information on this is here
We are currently working on various approaches to analysing data in OpenEvsys - which is quite important, considering the amount of information OpenEvsys can hold at any given time - which brings us to the whole analysis equation and the most effective way to achieve this and visualize the results. Patrick Ball's 'who did what to whom' model is certainly something to look at - I'd be interested in getting your advice on how best to achieve this. I'm also currently looking at the geo aspects and how GIS can be incorporated... more on this in the next post...
This reminds me of conversations with Robert Kirkpatrick a couple years ago on the need to identify Minimum Essential Indicators (MEI's) for decision-making. We can't collect everything, obviously, so we should make sure to understand what decision-makers base their decisions on.
Its possible now to do mapping quite cheaply with minimal infrastructure, though it can of course also get quite expensive. I think the first thing someone interested in mapping must ask themselves is "what do I want to do and why?" At this point, there are dozens, perhaps hundreds, of tools and software packages available, they are falling from the sky at this point. They will continue to improve and some advancements will be made, but there are a great many tools available now that we can all make use of. So, I would say that anything is possible, but the human rights group must define what it wants to do at the start. Note that that definition might be as simple as "we want to make a map showing where violations are taking place." Once that map has been produced, other things can be done with its data, like showing violations over time (ie as military units arrive), targeting satellites, or other tasks.
For equipment, technology, support, and expertise, that again depends on what you want to do. I think a fairly standard computer is enough for most mapping nowadays, and an internet connection is also important, especially to take advantage of google maps/earth etc. Its also probably important to have someone, either a staff or volunteer, who actually enjoys computing and making maps and stuff like that. Some of us just love this work, and to other people its totally frustrating and annoying. So, you want someone who loves it. For expertise, local people are the greatest experts on an area, whether you want to talk about mapping or human rights issues. Beyond that, you can contact me (lbromley 'at' aaas 'dot' org) or submit a request for GIS support at GISCorps.org.
One common problem people have is that they might know the area they
want to map, but don't know the latitude and longitude because they
cannot use GPS for various reasons (costs, risk, etc). Generally, the
biggest challenge of my group in mapping human rights violations in
Burma, Darfur, Ethiopia, and elsewhere is converting place names to
latitude and longitude. One problem is that how place names are
communicated might not match how they are spelled in databases, thus we
use "fuzzy" matching to help with this. Various tools can help with
this:
So, the above gives people a way to learn latitude and longitude for
locations of interest and then map those locations on Google Earth or
Google Maps. And you have your first map!
I'm interested in hearing and learning more about creative ways we can represent the rich data we collect on human rights violations or other conflict/crisis events. Conventional 2d maps are only one possibility. What might be some others?
I collected lots of historical data on battles and massacres in the Angolan civil war... but when I placed the 10,000 events on a 2d map, I found I didn't actually see very much compared with the rich stories/history and explanations underneath. In other words, the representation using a 2d static map only scratches the surface of the rich patterns lying beneath.
The stories I collected, the ways people understand an event, leveraging history or other local explanations, the ways in which reporting changes over time, and the networks of relationships between people on the ground, are still not well represented. I am convinced that in order to understand why events happen the way that they do, we need to find ways to better represent the attributes of each event (beyond the date and location).
Other than creating moving/animated maps that bring in the temporal dimension and using color and sound to bring further dimensionality to our representations, maybe we could brainstorm other ways people in the future might be able to interact with and understand patterns in the data...
What direction should we be moving in the area of understanding these data and creative modes of representation, given rapid technological advances that will enhance our ability to do just that? And how can we empower our students and the population at large to begin to ask good questions in this realm?
Thoughts?
Recently I thought creating a virtual reality/second life "game" in which people could interact in 3 dimensions with the data and take different "slices" from any direction, according to the question being asked, would be instructive. I look forward to hearing what we can come up with and brainstorming together with you!
This is an interesting issue that I've largely stayed away from in recent years but do have some thoughts on. Basically, the rule of thumb for most human rights work seems to be 'keep it simple' and while we can do complex visualizations, many groups worry that such viz would overwhelm or scare away viewers. Likewise, the evidentiary requirements I mention above are important, and visualization doesn't seem to help there. Lastly, I guess I am lucky that the imagery analysis that constitutes the bulk of our work is inherently visual and we don't need to jazz it up at all.
What I think is more applicable is visualization as an aid to analysis, which is a broader topic. Regardless, again, it seems simplicity is the key, though hard core analysts should have access to raw data to do their own work on. Also, it seems visualization, one way or another, is expensive, either in staff time or tools or both. With funds always tight, I haven't quite been able to dedicate lots of effort to the area.
I do note the 'GeoTime' application mentioned at the recent crisis mapper conference is nifty, but again I don't know how I'd apply it right now.
What I think is more applicable is visualization as an aid to analysis, which is a broader topic. Regardless, again, it seems simplicity is the key, though hard core analysts should have access to raw data to do their own work on.
When I think about data sets and the way we use them, simplicity is the answer for most of the users. What we want to do however is figure out ways to allow advanced users to play with simple tools that give them complex and otherwise hard to represent insights.
Do we have any concrete answers to what this might look like yet? Not at Ushahidi, though we're putting a lot of work into visualizations of data and about 20% of that time is spent on doing "crazy stuff that might not work", as we think this is where the breakthroughs might come. In a conversation with Patrick Meier a couple weeks ago we talked about applying algorithms from completely different fields to the Ushahidi data. Most of it probably won't work, but you never know until you experiment.
Closely aligned to this is the fact that I like to spend a lot of time on data visualiztion sites and blogs. There's always a fresh idea there, and the trick is figuring out how to apply it programmatically to the tools we build.
Patrick Ball's model for analyzing patterns and trends
Hi,
While documenting human rights violations, i think by following the Who?Did What? to Whom? When? and Where? model developed by Dr.Ball will allow us to analyse the pattern and trends of the violations more accurately. it will be more valuable if it is collected and supported by hard copy evidences such as statements, video clippings, 3d mapping, etc.
I'm interested in hearing and learning more about creative ways we can represent the rich data we collect on human rights violations or other conflict/crisis events. Conventional 2d maps are only one possibility. What might be some others?
I collected lots of historical data on battles and massacres in the Angolan civil war... but when I placed the 10,000 events on a 2d map, I found I didn't actually see very much compared with the rich stories/history and explanations underneath. In other words, the representation using a 2d static map only scratches the surface of the rich patterns lying beneath.
The stories I collected, the ways people understand an event, leveraging history or other local explanations, the ways in which reporting changes over time, and the networks of relationships between people on the ground, are still not well represented. I am convinced that in order to understand why events happen the way that they do, we need to find ways to better represent the attributes of each event (beyond the date and location).
Other than creating moving/animated maps that bring in the temporal dimension and using color and sound to bring further dimensionality to our representations, maybe we could brainstorm other ways people in the future might be able to interact with and understand patterns in the data...
What direction should we be moving in the area of understanding these data and creative modes of representation, given rapid technological advances that will enhance our ability to do just that? And how can we empower our students and the population at large to begin to ask good questions in this realm?
Thoughts?
Recently I thought creating a virtual reality/second life "game" in which people could interact in 3 dimensions with the data and take different "slices" from any direction, according to the question being asked, would be instructive. I look forward to hearing what we can come up with and brainstorming together with you!
Our Rainbow Europe map project is a real success, here is a link to a Brussels-based paper New Europe specialising in EU
affairs which has a special focus feature on LGBT rights in the EU and used our map to highlight the situation for LGBT people in Europe in one of their latest editions: http://www.neurope.eu/images/issues/858.pdf
I recently came across your blog and have been reading along. I thought I would leave my first comment. I don’t know what to say except that I have enjoyed reading. Nice blog. I will keep visiting this blog very often.
Yann,
I really appreciate that you raised this very important point right at the outset, "Some minimal requirements for geo-mapping the end user should know what he/she would like to see on the map." We can't get to where we're going if we don't know where we want to go. In human rights work - that requires us to have a VISION. Can we even imagine the way we want the world to be? And if we can, what information do we need to help us get there?
You also made a great point about the need to have a system - and the idea of developing a common data set for human rights violations. In January 2010, New Tactics will host a dialogue featuring databases for human rights advocacy and work. I would like to see how that discussion can link back to geo-mapping and your idea of developing a common data set.
There are a couple of longstanding human rights databases:
I'm very interested in seeing how such human rights data systems can be utilized to render geo-maps (taking into consideration the critical points raised already in this dialogue, especially ensuring that people are not further endangered or their families/communities compromised.)
Nancy Pearson, New Tactics in Human Rights Program Manager
OpenEvsys is a human rights case management system, based on the Event Standard Format. The system was built for HURIDOCS as per their requirements, by Respere. OpenEvsys is a Free Open Source project licenced under AGPL v3 - we are currently in the process of building the community around this. More detailed information on this is here
We are currently working on various approaches to analysing data in OpenEvsys - which is quite important, considering the amount of information OpenEvsys can hold at any given time - which brings us to the whole analysis equation and the most effective way to achieve this and visualize the results. Patrick Ball's 'who did what to whom' model is certainly something to look at - I'd be interested in getting your advice on how best to achieve this. I'm also currently looking at the geo aspects and how GIS can be incorporated... more on this in the next post...
Mifan Careem - Respere, Sahana, OpenEvsys
Hi Patrick,
Definitely something to look at... I'm sure the outcome would be great.
Mifan Careem - Respere, Sahana, OpenEvsys
This reminds me of conversations with Robert Kirkpatrick a couple years ago on the need to identify Minimum Essential Indicators (MEI's) for decision-making. We can't collect everything, obviously, so we should make sure to understand what decision-makers base their decisions on.
Patrick Meier, Ushahidi
Its possible now to do mapping quite cheaply with minimal infrastructure, though it can of course also get quite expensive. I think the first thing someone interested in mapping must ask themselves is "what do I want to do and why?" At this point, there are dozens, perhaps hundreds, of tools and software packages available, they are falling from the sky at this point. They will continue to improve and some advancements will be made, but there are a great many tools available now that we can all make use of. So, I would say that anything is possible, but the human rights group must define what it wants to do at the start. Note that that definition might be as simple as "we want to make a map showing where violations are taking place." Once that map has been produced, other things can be done with its data, like showing violations over time (ie as military units arrive), targeting satellites, or other tasks.
For equipment, technology, support, and expertise, that again depends on what you want to do. I think a fairly standard computer is enough for most mapping nowadays, and an internet connection is also important, especially to take advantage of google maps/earth etc. Its also probably important to have someone, either a staff or volunteer, who actually enjoys computing and making maps and stuff like that. Some of us just love this work, and to other people its totally frustrating and annoying. So, you want someone who loves it. For expertise, local people are the greatest experts on an area, whether you want to talk about mapping or human rights issues. Beyond that, you can contact me (lbromley 'at' aaas 'dot' org) or submit a request for GIS support at GISCorps.org.
From on of my other posts here, this is a simple set of links to help mapping. A simple way to create Keyhole Markup Language (KML), the basis for Google Maps and Google Earth, is to use the Google Docs spreadsheet mapper (http://earth.google.com/outreach/tutorial_spreadsheet.html)
to enter your information and latitude and longitude to create a KML.
If you don't want to use the Google Docs system, use one of the
standalone Excel KML makers, such as http://www.earthpoint.us/ExcelToKml.aspx or http://freegeographytools.com/2007/xls2kml-another-excel-to-kml-converte... .
One common problem people have is that they might know the area they
want to map, but don't know the latitude and longitude because they
cannot use GPS for various reasons (costs, risk, etc). Generally, the
biggest challenge of my group in mapping human rights violations in
Burma, Darfur, Ethiopia, and elsewhere is converting place names to
latitude and longitude. One problem is that how place names are
communicated might not match how they are spelled in databases, thus we
use "fuzzy" matching to help with this. Various tools can help with
this:
1) The AAAS fuzzy matchers for Burma, Darfur, Ethiopia, and Pakistan, at: http://sustsci.aaas.org/quickmatch/
2) Wider area fuzzy matcher from the EU at: http://dma.jrc.it/services/fuzzyg/
3) Another wide area fuzzy matcher from the US Dept of Defense: http://earth-info.nga.mil/gns/html/index.html
So, the above gives people a way to learn latitude and longitude for
locations of interest and then map those locations on Google Earth or
Google Maps. And you have your first map!
Lars Bromley, AAAS-SHRP
Hi Everyone!
I'm interested in hearing and learning more about creative ways we can represent the rich data we collect on human rights violations or other conflict/crisis events. Conventional 2d maps are only one possibility. What might be some others?
I collected lots of historical data on battles and massacres in the Angolan civil war... but when I placed the 10,000 events on a 2d map, I found I didn't actually see very much compared with the rich stories/history and explanations underneath. In other words, the representation using a 2d static map only scratches the surface of the rich patterns lying beneath.
The stories I collected, the ways people understand an event, leveraging history or other local explanations, the ways in which reporting changes over time, and the networks of relationships between people on the ground, are still not well represented. I am convinced that in order to understand why events happen the way that they do, we need to find ways to better represent the attributes of each event (beyond the date and location).
Other than creating moving/animated maps that bring in the temporal dimension and using color and sound to bring further dimensionality to our representations, maybe we could brainstorm other ways people in the future might be able to interact with and understand patterns in the data...
What direction should we be moving in the area of understanding these data and creative modes of representation, given rapid technological advances that will enhance our ability to do just that? And how can we empower our students and the population at large to begin to ask good questions in this realm?
Thoughts?
Recently I thought creating a virtual reality/second life "game" in which people could interact in 3 dimensions with the data and take different "slices" from any direction, according to the question being asked, would be instructive. I look forward to hearing what we can come up with and brainstorming together with you!
This is an interesting issue that I've largely stayed away from in recent years but do have some thoughts on. Basically, the rule of thumb for most human rights work seems to be 'keep it simple' and while we can do complex visualizations, many groups worry that such viz would overwhelm or scare away viewers. Likewise, the evidentiary requirements I mention above are important, and visualization doesn't seem to help there. Lastly, I guess I am lucky that the imagery analysis that constitutes the bulk of our work is inherently visual and we don't need to jazz it up at all.
What I think is more applicable is visualization as an aid to analysis, which is a broader topic. Regardless, again, it seems simplicity is the key, though hard core analysts should have access to raw data to do their own work on. Also, it seems visualization, one way or another, is expensive, either in staff time or tools or both. With funds always tight, I haven't quite been able to dedicate lots of effort to the area.
I do note the 'GeoTime' application mentioned at the recent crisis mapper conference is nifty, but again I don't know how I'd apply it right now.
Im with lbromley on this:
When I think about data sets and the way we use them, simplicity is the answer for most of the users. What we want to do however is figure out ways to allow advanced users to play with simple tools that give them complex and otherwise hard to represent insights.
Do we have any concrete answers to what this might look like yet? Not at Ushahidi, though we're putting a lot of work into visualizations of data and about 20% of that time is spent on doing "crazy stuff that might not work", as we think this is where the breakthroughs might come. In a conversation with Patrick Meier a couple weeks ago we talked about applying algorithms from completely different fields to the Ushahidi data. Most of it probably won't work, but you never know until you experiment.
Closely aligned to this is the fact that I like to spend a lot of time on data visualiztion sites and blogs. There's always a fresh idea there, and the trick is figuring out how to apply it programmatically to the tools we build.
Hi,
While documenting human rights violations, i think by following the Who?Did What? to Whom? When? and Where? model developed by Dr.Ball will allow us to analyse the pattern and trends of the violations more accurately. it will be more valuable if it is collected and supported by hard copy evidences such as statements, video clippings, 3d mapping, etc.
Hi Everyone!
I'm interested in hearing and learning more about creative ways we can represent the rich data we collect on human rights violations or other conflict/crisis events. Conventional 2d maps are only one possibility. What might be some others?
I collected lots of historical data on battles and massacres in the Angolan civil war... but when I placed the 10,000 events on a 2d map, I found I didn't actually see very much compared with the rich stories/history and explanations underneath. In other words, the representation using a 2d static map only scratches the surface of the rich patterns lying beneath.
The stories I collected, the ways people understand an event, leveraging history or other local explanations, the ways in which reporting changes over time, and the networks of relationships between people on the ground, are still not well represented. I am convinced that in order to understand why events happen the way that they do, we need to find ways to better represent the attributes of each event (beyond the date and location).
Other than creating moving/animated maps that bring in the temporal dimension and using color and sound to bring further dimensionality to our representations, maybe we could brainstorm other ways people in the future might be able to interact with and understand patterns in the data...
What direction should we be moving in the area of understanding these data and creative modes of representation, given rapid technological advances that will enhance our ability to do just that? And how can we empower our students and the population at large to begin to ask good questions in this realm?
Thoughts?
Recently I thought creating a virtual reality/second life "game" in which people could interact in 3 dimensions with the data and take different "slices" from any direction, according to the question being asked, would be instructive. I look forward to hearing what we can come up with and brainstorming together with you!
Our Rainbow Europe map project is a real success, here is a link to a Brussels-based paper New Europe specialising in EU
affairs which has a special focus feature on LGBT rights in the EU and used our map to highlight the situation for LGBT people in Europe in one of their latest editions: http://www.neurope.eu/images/issues/858.pdf
Please see pages 8-9
I recently came across your blog and have been reading along. I thought I would leave my first comment. I don’t know what to say except that I have enjoyed reading. Nice blog. I will keep visiting this blog very often.
Mcford University | Mcford University | Mcford University
Pages