Lessons from Ushahidi
Juliana Rotich, co-founder Ushahidi
What is Ushahidi? <a href="http://ushahidi.com">Ushahidi</a>, which means 'testimony' in Swahili, is a platform that was initially developed to map reports of violence in Kenya after the post-election fallout at the beginning of 2008.
The Ushahidi Story: The volunteer team behind Ushahidi rapidly developed a tool for Kenyans to report and map incidents of violence that they saw via SMS, email or the web. Within a week Ushahidi had gone from idea to live deployment and to live online prototype. Ushahidi now serves as a prototype and a lesson for what can be done by combining crisis information from citizen- generated reports, media and NGOs and mashing that data up with geographical mapping tools. The Ushahidi platform has been used to monitor elections in India, Mexico, Lebanon and Afghanistan. It has been deployed in the DR Congo to track unrest, and Zambia to monitor medicine stockouts.
1. Lessons learned from the first iteration of the platform in Kenya
In the paper Innovation for Africa, Christopher Fabian and Erica Kochi mention the idea of varying shades of connectivity, to describe how various populations may not have access to the internet, but they do have access to mobile phones. Ushahidi provides a way to aggregate information from all areas in the spectrum of connectivity. In the same vein, this capability of aggregation also depends on the general technological landscape of the country where the implementation of Ushahidi is done. In the paper I co-wrote with Joshua Goldstein, we looked <a href="http://cyber.law.harvard.edu/publications/2008/Digitally_Networked_Technology_Kenyas_Post-Election_Crisis">the role of digitally networked technology in Kenya's 2007-2008 post election crisis</a>. In the paper, we describe three important ways that Kenyans used new technology to coordinate action: SMS campaigns to promote violence, blogs to challenge mainstream media narratives, and online campaigns to promote awareness of human rights violations. The Kenyan case hints at the interconnectedness of these modes of information dissemination, with an important component being that of a vibrant and connected blogosphere.
Essentially, where is the crowd most active? What are the nodes on the network and who are the key influencers in said network? Ushahidi emerged out of the connections made online between bloggers, the solidarity they felt, and the actions they took to contribute positively during a difficult time. First, the bloggers were in touch via SMS, sharing information with each other, posting on their blogs, posting pictures on flickr and twittering. Blogposts were then aggregated on the popular Kenyan aggregator <a href=", http://www.kenyaunlimited.com/kenyan-blogs-webring/">Kenyaunlimited</a> and<a href="http://www.globalvoicesonline.org/2007/05/10/kenya-blogging-the-next-presidential-election-post-by-post/"> Globalvoices online</a>. The information on Global Voices was picked up by connected individuals such as <a href="http://ethanzuckerman.com/blog">Ethan Zuckerman</a>, a noted academic and Afrophile. With his amplification, the crisis got a context that goes beyond the clips on CNN or other international media. Successful implementations not only depend on the offline strategy for letting people know how to report, but the technological landscape of the country in question. It does not necessarily have to be a wired country, at the time Ushahidi launched, less than 5 percent of Kenyans had regular Internet access. (Internet penetration is difficult to measure in Africa, where the majority of Internet users have shared access, primarily through Internet cafes. While publicly available figures in 2008 put access at 1.5%, this only refers to broadband subscriptions, which account for a tiny percentage of access. With the advent of fiber optic links in 2009, the tech landscape in Kenya is changing) What is instructive here is the role of transnational activists and diaspora in online platforms. Engaging these activists and attendant diaspora is very useful when implementing Ushahidi for advocacy. Evaluating the technological landscape, from wireless network coverage, state of the blogosphere, and interconnectedness of the same is an important lesson that can be gained from the Ushahidi experience in Kenya.
Ushahidi went live within four days. Just as Google encourages rapid prototyping, there is value in remembering that if an organization is thinking of crowd-sourcing, they do not have to start from scratch. The technology is provided as free and open source software making it relatively easy to be up and running in no time. The installation process has been simplified, and there are plans to offer a complete hosting solution. This would work similar to the wordpress blog sign up process, which is as easy as setting up a new email account. This will take time to achieve, but meanwhile, interested individuals with basic IT knowledge can set up an instance easily. They should just <a href="http://download.ushahidi.com">download the software</a> and as famous Nike ad says... "Just do it!"
2. Strategic Considerations
After Ushahidi received major funding in mid 2008, we started working with several NGO's (Non Governmental Organizations) in Kenya to test the Alpha version of the software. One of those organizations was<a href="www.mediafocusonafrica.org"> Media Focus on Africa (MFOA)</a>. In Conjunction with<a href="www.butterflyworks.org"> Butterfly works</a>, Ushahidi provided the platform and hosting space for a campaign that went live on November 21st 2008. <a href="http://peaceheroes.ushahidi.com">Unsung Peace Heroes</a> used Ushahidi to collect nominations, post the nominations and map the locations of the peace efforts. Nominators could send nominations of unsung peace via the Peace Heroes Ushahidi site, via SMS and email, and by filling out a paper nomination at various peace events. The <a href="http://wiki.ushahidi.com/doku.php?id=case_studies">use case</a> is instructive in that it provides an important lesson on strategy. I.e, the tools or the platform is not the only thing organizations need to have in place for a successful implementation. The online, mobile and offline strategy is also important. Thinking about how activities on each of those realms work together is very important, because the Ushahidi platform aggregates outputs of those activities. How is the information being gathered? Are people encouraged to submit information? If so, are the instructions very clear on what information the crowd should provide? More importantly, does the person submitting information get confirmation that their message has been received? The feedback loop is one thing to keep in mind.
Marten Schoonman, the coordinator of the Peace Heroes project noted:
“It is a nomination process. Representation of what is going on in an area is helpful.” However, he also indicated that there is one major drawback of an Internet-based project: “The people who MFOA is targeting do not have Internet access - We want to bring the results back to the people using mass media.” As a result of MFOA’s goals, Internet is only a part of their larger projects and campaigns. For the Unsung Peace Heroes project, MFOA used a multimedia approach, including an online presence, newspaper ads, radio and television appearances, participation in live events and word-of-mouth. For this campaign, the site enabled “MFOA to select winners from various parts of the country more easily.”
According to Marten, “the distribution of fliers by hand during relevant events (peace in this case) worked wonders.” He believes “that people participate with a certain mindset for that day and find it attractive to participate and spend a few shillings.” (in sending the text msg to nominate someone).
3. What lessons have we learned so far from the application of tools for crowdsourcing information for crises mapping?
When mapping crises, there is the 1% rule to keep in mind. A majority of the reports received during the height of a crisis may come from an active group of participants who will provide the bulk of information.In a dynamic situation, that percentage could increase, and a small trickle of information could quickly become a flood. In this interconnected online world, it is important to know how to sift through the information, isolating the signal from the noise. Ushahidi does not have an answer to this, but with the first step of visualizing incoming data based on geographical location, it becomes easier to figure out what is happening where.
In the legacy version of Ushahidi, 220 incident reports were received on the website, submitted by 131 unique contributors (as measured using unique IP addresses). It is a worthy goal to analyze the data that your implementations receives, with an eye towards diversity of sources. With the integration of twitter in the Ushahidi platform, we are learning from events such as the Iranian Elections and the <a href="http://www.ethanzuckerman.com/blog/2009/04/13/studying-twitter-and-the-moldovan-protests/">Moldovan protests</a>, to approach integration of twitter information carefully as many reports that might be aggregated into an Ushahidi instance may be retweets (a message that is resent by multiple people) that may suggest a flash point, though that may not necessarily be the case. To this end, the Ushahidi platform is continuing work on the Swift River initiative.
Swift river is an initiative that came out of the Ushahidi community that seeks to do two very important things, both of which are crucial for not just Ushahidi, but for many emergency response activities in the future. First, it gathers as many possible streams of data about a particular crisis event as possible. Second, using a two-part filter, that stream of data is filtered through both machine based algorithms and humans to better understand the veracity and level of importance of any piece of information.
4. Lessons learned from other implementations such as votereport.in (Elections and social engagement) by organizations working independently or in conjunction with Ushahidi
Just as there are varying shades of connectivity, there are varying needs for localization. Organizations using the platform for election monitoring such as Alive In Afghanistan greatly benefit from the design elements developed for previous implementation like votereport.in. Ushahidi is currently working with <a href="http://emoksha.org">Emoksha.org</a> and <a href="http://meedan.net">Meedan</a> to create a template for others to use a flavor of Ushahidi specifically tailored to election monitoring.
Localization of software is a very important consideration when implementing Ushahidi. We learned this first hand with the deployment of Ushahidi in the Democratic Republic of Congo. The lingua franca there is French, and Ushahidi quickly adapted the platform to include the translation of French to English and vice versa. The localization efforts (translation of code to Spanish) by Cuidemos El Voto have helped Ushahidi provide support to testers interested in deploying Ushahidi in Central America. These efforts continue with the translation framework provided by use of Pootle Software to power <a href="http://translate.ushahidi.com">http://translate.ushahidi.com</a>
Having a network or partners on the ground is also a very important lesson that we learned from the DRC implementation and also from working with <a href="http://votereport.in">votereport.in</a>. Partnering with civil society, media and bloggers on the ground helps to get the word out on the implementation, and it also provides a network which the administrator of a platform can use to confirm veracity of reports received.
5. Challenges. For instance, problems that you've had in working with mobile service operators and in authenticating information provided by the users.
Various parts of the world are mapped to varying levels of accuracy. The platform provides options to select mapping providers, and this helps to deal with this challenge, but in completely unmapped areas, the challenge of 'pinning data' on the map remains. Where do you pin information to if you do not know the latitude and longitude or the given name for a certain area? An interesting project that provides an example to learn from is the <a href="http://mapkibera.org">Map Kibera project</a>. The project embarked on a mapping endeavor that created the first public digital map of Kibera, one of Africa's largest slums. Mikel Maron of the Humanitarian OpenStreetMap Team worked with volunteers, techies, and other partners to train youth on current mapping techniques. Over a two week period in November, the team made a once unmapped area visible in the digital domain. The techniques they used such as walking papers and use of GPS (Global Positioning Systems) are worth looking into and replicating for unmapped areas.
Data poisoning is another challenge for implementations of Ushahidi. If an organization uses the platform for advocacy, how would they deal with the influx of information from antagonists or even organized government interference, something Evgeny Morozov of Foreign Policy's Net Effect calls <a href="http://neteffect.foreignpolicy.com/posts/2009/07/15/the_spinternet_goes_global_the_nigerian_chapter">'The Spinternet'</a>? This is a concern of many advocacy organizations. There is no straight solution for this other than openness and showing clearly on each report that the information is either verified or not. The platform is built to take into account participation (ability to comment on existing reports). Ushahidi saw participation from the public in helping correct some reported incidents, very similar to how Wikipedia functions.
Verification of received reports is often difficult, how can you tell that the information submitted is indeed true? The platform currently provides an administrator pane that presents incoming information for review before it is 'published' for view on the website. It is still possible to get some level of verification with hyperlocal involvement with NGO’s and partners on the ground. The other issue that is the bane of most online platforms is spam, both in the comment system that adds onto existing reports on the mapping, but there is also the possibility of received SMS spam. These issues have been addressed by providing a spam filter as part of the platform.
One of the big questions that Ushahidi had to wrestle with was "What is in it for a person contributing reports?" Why should they submit information, and what could use could an implementation be? The platform provides a feedback loop to help organizations answer this question in their own implementations. The loop is closed with the alerts system whereby an implementation can be configured such that a user can specify whether they would like to receive and email, SMS or an RSS (Really Simple Syndication) alerts when a new report is submitted in a certain location. With implementations such as <a href="http://hatari.co.ke">hatari.co.ke</a>, Ushahidi is seeking to understand how to fine tune the feedback loop in order to provide value to users of the platform.
With SMS as part of the platform, it introduces the problem of cost for managing an active campaign. It is difficult to get cost effective bulk SMS rates and short codes for NGOs and some organizations have to opt for only email alerts as part of their implementation. Some organizations have partnerships with mobile operators, but their systems for use of the SMS allocations do not provide for API (Application Programming Interfaces) that can be utilized by the platform. What we've learned is that providing a way to sync information using available open source programs such as <a href="http://frontlinesms.com">FrontlineSMS</a> that work by turning a laptop and a mobile phone into a low cost communications hub is more expedient. Partnership with the creators of this software will yield ways for NGOs to run campaigns within their means and to use available hardware to get information out, or to collect data. They need not wait on approval of a proposal by the CSR (corporate social responsibility) arm of a major mobile service provider.
6. Successes
In less than two years, the Ushahidi platform has grown from just an idea to a global community of programmers and volunteers. This is perhaps the greatest success. With more than 500 downloads of the software, there are bound to be more innovative implementations that build on the current platform. As the core platform is improves with each iteration, the tool will be of more utility to those interested in crowd-sourcing crisis information. The release of mobile applications for Java enabled phones, Android, Windows Mobile and Iphone goes a long way in ensuring that information can be gathered in myriad ways and aggregated online.
Juliana Rotich is Co-founder and the Program Director of Ushahidi. Ushahidi (which means “testimony” in Swahili), a web application created to map the reported incidents of violence during the post-election crisis in Kenya. Currently, she is working with a team to continue development of this new free and open source platform that makes it easier to crowd-source crisis information and visualize data. [[Category:{{{Category}}}]]