Infocrisis.social: A Socially-powered Control Panel

We present InfoCrisis.Social, a platform that aggregates crisis related information to help individuals and communities to develop situational awareness in a country continuously exposed to natural threats. The platform combines governmental and mass media reports as well as real time social media information sources to provide communities with the most relevant info at each stage in a potential or actual disaster, generating lists, graphs and maps that allow users to make sense of such chaotic situations. The assumption that informs our citizen crowdsourcing project is based on the idea that communication and information are core in a crisis response. Information per se helps those at risk to take action in reducing the damage caused by natural hazards and participating in the distribution of vital information can help mitigate the psychosocial trauma elicited during a crisis.

This presentation will cover the discovery and design aspects of our UX process, we have conducted a series of co-design workshops and user feedback sessions. This project has also involved some compelling dimensions of information architecture such as the normalization of information from multiple sources. We will tell the story of the project from initial sketches and concept models, subsequent design workshops, the wireframing stages, and the development of prototypes.

Transcript

I work at Cigiden, a consortium of six universities and public-funded research institute. We’ve partnered for this project with Inria, a French research center focused on math and computer science. They have an office here in Chile which also has public support and funding. We’ve joined forces to build this platform called InfoCrisis.Social, which in English would be the other way around: social crisis information. That makes a little bit more sense.

As Jen was mentioning, Chile is a country marked by disaster. We have this timeline that my office has developed. In 1960, we had the strongest earthquake recorded in history. We have multiple types of risk. Not only earthquakes. We also have drought, active volcanoes, snowstorms, floods, and wildfires. In 2010, we had a massive earthquake in Maule. A rupture from the earth was 350 kilometers long and it lasted for over four minutes with a strength of over 8 milliwatts. Picture that it’s shaking so strongly you can’t stand on your feet. It went for four minutes. It was so intense that it was felt in Sao Paulo, Brazil. After that, came a huge tsunami that washed away a few coastal towns. Even this year, we’ve already had a couple of important disasters. Ending last year on Christmas day, was a huge earthquake. But we have good infrastructure for this. Fortunately it happened in a place in the country that, although it’s very rural and vulnerable, everything is built in wood, so the houses had a lot of strength to resist this. But by the turn of the new year, a huge forest fire started coming into a city. Between 100 and 200 houses were lost in that. Then, just a couple of weeks after that, another huge cluster of forest fires started in central south Chile. It wouldn’t stop. We had a huge heat wave and extended drought over many years, so the fires weren’t stopping. We had help from many parts of the world. Countries sent their hugest firefighting airplanes. In that picture there’s a 747. The major capital was covered in smoke from fires that were happening miles away from it. It was only in the end, it was only doused by some unusual rain. We were lucky for that. Basically, we’re used to living with disasters of many kinds. Our country has always had to start rebuilding itself from the ground up. What we want to do is help people be informed in these situations.

Another interesting thing about our country is that we’re very well-connected. Social media adoption is very quick and very intense. People are very good at chatting on social media. Very dense connections and relationships get covered through multiple media. We have the third rate of penetration for social media in the world. Of course, when disasters come, people come to social media to talk about them. We even laugh about that. The tweet you see here at the corner, they were playing with these hashtags they built to have a game rolling. They’re talking about how before the internet, you can report when you had an earthquake, but now you’re so used to saying, “Oh, it’s quaking.” We just make fun of it. We also see here “forest fires” are the top trending topic. It’s something that’s happening these days. Last week we were very concerned about that. People sent all kinds of information. Maps of affected areas and firsthand testimonies of guys saying, “We put away six fires today and unfortunately they were all intentional.” So a very active country on social media. We’re going to use this participation we have, this opportunity. There’s rich information going on about the situation but there’s no central point that can gather this information to help people make sense of what’s called in disaster management “situational awareness”, of being really aware of how far it is, how big it is, etc.

That brings us to our project. Our project is focused on providing citizens, reporters, and researchers with real-time information information related to these natural disasters and crises. We want to be a hub to centralize different sources of information. The project started out as a sim. The first steps in terms of our research was to talk to people in our lab. We have these experts in disaster from many areas. We spoke with our researchers and students. I did a quick poll of the most important types of information that you would need for different types of disasters, depending on who you are. Here we focused on what type of information we should provide on this dashboard.

With that, we did the first round of our conceptual model. We’re scraping some official websites to get information. We do some stuff with Twitter an different ways of manipulating the tweets, forming different feeds from that. From that, we can also create analytics and charts. We also integrate the different maps that we have, as seen in the green circles at the bottom. As well as other types of information. The idea is that all of this information is gathered in this dashboard and the users can rate the resources and suggest additional resources. We want to make this a very participative system. We want to build a community around this disaster topic. We know there’s a Latin community on Twitter. There are a lot of people who are focused on informing disasters. We want to make this participation something that’s a little bit more structured, but still give it enough freedom.

Another interesting thing about this project is how we’ve been organizing information. Depending on the stage of disaster, different information becomes a different priority. When you’re in a very threatening situation and there’s an evacuation in place, you just need to see evacuation maps. Our priority will be to put up educational information so people will be prepared on how they can connected with their families and how they can have emergency kits, backup batteries for their phones, etc.

Here we have some initial wireframes we did for this project. We started mapping out how we wanted to integrate maps with official alerts and different news sources, like videos. We wanted incentives for participation for people on Twitter. We wanted to put up the most relevant hashtags and user accounts. We need to figure out how to avoid people gaming the system.

Then we started this partnership with Inria. Here are a couple of pictures from our meetings. Since then, we’ve been doing a lot of brainstorming. We came up with a name and logo for the project, among other stuff. We did this development-oriented conceptual model, where we show that users see different types of information. Dynamic reports are collected. How different types of disasters have different sources and the disaster will show us different priorities. Then we also prepared a poster for our conference where we did a public-facing explanation of the project. We have information from official sources, from the press, and from people on social media. We can create maps. We can capture tweets. Facebook videos and pictures. What you see here at the bottom is the users who are seeing these resources can evaluate them and share them. We want to promote people to be active on this, without a high bar for participation. We want to make it very easy for them to participate. This feedback will help us train the system with a combination of artificial intelligence and user curation. We’ve also been doing content modeling. It was very tricky to figure out what an alert means. Different institutions have different levels for alerts. We’re trying to shape information around the stage and type of event. We have a very important factor in terms of location. Geolocation of information will be a very critical part of this project. With all of this, we’ve been working on content inventories to determine information sources. Different institutions: NGOs, media, and, of course, digital volunteers. These are active members of the community who are already focused on the topic of disasters. We want to make them our allies and to be a central part of this community. And we collect information from different channels that we have here on the right.

Another interesting discovery part of our research has been design workshops with different groups of people. We have disaster experts, computer experts, and we did some with social media, these digital media volunteers. These design workshops are where people discuss the types of important information they need in different contexts and hash out how they would see them on a screen. We did this focused on a few scenarios. In the first session, we had two groups, then we joined them all together. It’s led to more sketching. We’re working on these interfaces. Another part of our project is we want to avoid the problem we’re seeing from some of the sources that have great information that people need during a disaster but they don’t know how to show it in the most clear possible way. Some very institutionalized, kind of military organizations, they don’t know much about information design. This is just the first sketch.

Currently, we have a first release, based on our development timeline on Agile. Our initial development work has focused on capturing Twitter information. Here we’re showing that according to different types of disaster and information, based on the type of source that we have. Official sources, press, NGOs, digital volunteers, and community. As you can see here, from the last couple of days, there’s been a lot of activity, including a fire, earthquake, tsunami, and flooding.

So, how is this project sustainable? At the very core, our main goal is to build a resilient community. If people had less vulnerability, there’s less damage from nature hazards. Disaster will not have as much damage to infrastructure and there’s less rebuilding. A resilient community is more aware of their natural resources and they know what to protect and how to work with the risks. They will be able to live in better connection with their environment. This makes them less vulnerable to natural hazards and disasters. Then there’s the technical part. A central repository will have less traffic. We’re improving information delivery and news and software. We’re pretty focused on that. In terms of resources, there are some starting resources. We’ve also been working a lot with volunteers. That’s one of the things we want to bring to you. If anybody is interested in joining our project, we would be happy to speak with you. We’re very interested in getting people’s input.

Biography

Javier Velasco-Martin, PH.D. in Information Science, focuses on communication dynamics in ICTs. He was an early member of the global IA/UX community and pioneered the field in Latin America. With over 16 years of experience as UX researcher and designer in leading organizations in Chile and the US, he is currently postdoctoral researcher at CIGIDEN, a national research center focused on integrated disaster management.

@mantruc

Gonzalo Bacigalupe, EdD, MPH, is professor of Counseling Psychology at the University of Massachusetts Boston. He is principal investigator of the #ICT4D at the Research Center for Integrated Disaster Risk Management (CIGIDEN), senior invited faculty member in the doctoral program of the Catholic University of Chile in Valparaiso, and adjunct professor at the Catholic University of Chile School of Engineering. His research with colleagues in Chile, Mexico, Portugal, Spain, Turkey, and the USA, focuses on the impact of emerging media adoption on families, the role of patient online communities, the use of social technologies to build community resilience during socionatural disasters, transnational families, and family health.