Three reading suggestions on Social media and Crisis/Emergencies/Mass Disruptions #ISCRAM

The ISCRAM 2012 conference is now on its second day in Vancouver Canada. The conference has been great both in terms of good research and in terms of networking opportunities with fellow researchers. So far, I would like to present three papers that could be of interest to look a bit deeper into. These papers cover aspects of social media and how the public and government agencies use social media. The papers also cover various aspects on how to make use of insights in social media to support and improve response work. In a few days, all papers from the conference will be published on the website. Meanwhile, download, read and discuss the following papers with your friends at work.

[PDFLearning from the Crowd: Collaborative Filtering Techniques for Identifying On-the-Ground Twitterers during Mass Disruptions 
Kate Starbird, Grace Muzny, Leysia Palen

Social media tools, including the microblogging platform Twitter, have been appropriated during mass disruption events by those affected as well as the digitally-convergent crowd. Though tweets sent by those local to an event could be a resource both for responders and those affected, most Twitter activity during mass disruption events is generated by the remote crowd. Tweets from the remote crowd can be seen as noise that must be filtered, but another perspective considers crowd activity as a filtering and recommendation mechanism. This paper tests the hypothesis that crowd behavior can serve as a collaborative filter for identifying people tweeting from the ground during a mass disruption event. We test two models for classifying on-the-ground Twitterers, finding that machine learning techniques using a Support Vector Machine with asymmetric soft margins can be effective in identifying those likely to be on the ground during a mass disruption event.

[PDFConnected Communications: Network Structures of Official Communications in a Technological Disaster 
Jeannette N. Sutton, Britta Johnson, Mathew Greczek, Emma S. Spiro, Sean M. Fitzhugh, and Carter T. Butts

Informal online communication channels are being utilized for official communications in disaster contexts. Channels such as networked microblogging enable public officials to broadcast messages as well as engage in direct communication exchange with individuals. Here we investigate online information exchange behaviors of a set of state and federal organizations during the Deepwater Horizon 2010 oil spill disaster. Using data from the popular microblogging service Twitter, we analyze the roles individual organizations play in the dissemination of information to the general public online, and the conversational microstructure of official posts. We discuss characteristics and features of following networks, centrality, and conversational dynamics that may affect information exchange in disaster. This research provides insight into the use of networked communications during an event of heightened public concern, describes implications of conversational features, and suggests directions for future research.

[PDFTowards a realtime Twitter analysis during crises for operational crisis management 
Teun Terpstra, R. Stronkman, A. de Vries, G.L. Paradies

Today’ s  crises  attract  great  attention  on  social  media,  from  local  and  distant  citizens  as  well  as  from   news media. This study investigates the possibilities of real-time and automated analysis of Twitter messages during crises. The analysis was performed through application of an information extraction tool to nearly 97,000 tweets that were published shortly before, during and after a storm hit the Pukkelpop 2011 festival in Belgium. As soon as the storm hit the festival tweet activity increased exponentially, peaking at 576 tweets per minute. The extraction tool enabled analyzing tweets through predefined (geo)graphical displays, message content filters (damage, casualties) and tweet type filters (e.g., retweets). Important topics that emerged  were  ‘early  warning  tweets’,  ‘rumors’  and  the  ‘self- organization  of  disaster  relief’  on  Twitter .  Results  indicate  that  automated  filtering  of  information   provides valuable information for operational response and crisis communication. Steps for further research are discussed.

Samverkansoperatörer i gränsöverskridande operatörskluster

Igår hade jag förmånen att få genomföra observationer och informella intervjuer kring ett fenomen som traditionellt endast varit möjligt att studera inom ramen för kontrollerade experiment eller i samband med övning. Det jag studerade var operativ samverkan hos samlokaliserade ledningsoperatörer från de centrala blåljusaktörerna, i samband med riktiga insatser.

De olika ledningsoperatörerna med sina befintliga tekniska systemstöd var fysiskt samlokaliserade. Operatörerna satt med sina operatörsbord på mycket kort avstånd till varandra. Operatörerna hanterade skarpa larm med hjälp av sina befintliga tekniska system.

  • Polisen hade 1st LKC-operatör på plats
  • SOS-alarm hade 2st ambulansdirigenter på plats
  • De två räddningstjänstförbunden hade vars 1st ledningsoperatör på plats

Operatörerna ersatte inte de operatörer som fanns i de befintliga ledningscentralerna hos varje organisation utan formerade snarare ett ledningsoperatörskluster med egenskaper vi sällan kunnat studera i verkliga händelser. Operatörerna hade olika modell för hur deras samverkande roll länkades in sina organisationers befintliga bakomliggande ledningsstrukturer.

Detta “operatörskluster”, som dessutom saknade ett övergripande befäl eller annan form av formell styrning, visade sig kunna överbrygga och kompensera för de organisatoriska och tekniska brister som dagens stuprörsorganisering orsakar. Operatörerna arbetade som ett team med att säkra upp att information som fanns relaterat till pågående händelser snabbt förmedlades till de andra organisationerna.

Min korta studie som omfattade endast 5 timmar där dessa “samverkansoperatörer” påvisade många spännande och positiva egenskaper som verkar följa av fysiskt samlokaliserade operatörsteam. Följande händelser hanterades:
– Farligt gods / kem-händelse (explosivt-material)
– Suicid
– Trafikolyckor i tunnelmiljö
– Trafikolycka, tung bärgning
– Eftersök av saknad person, ev suicid
– Hjärtstoppslarm
– Brand i byggnad (lackverkstad)
– Omhändertagande av våldsam person

Analysen av studien är naturligtvis inte klar men gårdagens studie gav insikter om att det finns en rad nya och ibland oväntade former av samverkan som kan utveckla Svensk olycks- och krishantering.

Twitcident: Delft University takes the lead in making use of social media for incident response

Delft University is now taking the lead in innovative design when it comes to making use of social media for incident response. A group of researchers with Fabian Abel, Claudia Hauff, Geert-Jan Houben, Richard Stronkman and Ke Tao focusing on Social Semantic Web research have now launched a great innovation called Twitcident, where they harvest social media insights after an incident has happened. Read more about the research at:

Additional information could also be found at:
Blog: (in Dutch)
Twitter: @twitcidentapp

  • Fabian Abel, Claudia Hauff, Geert-Jan Houben, Richard Stronkman, Ke Tao. Semantics + Filtering + Search = Twitcident. Exploring Information in Social Web Streams. In Proceedings of International Conference on Hypertext and Social Media (Hypertext), Milwaukee, USA, 2012 [pdf]

In relation to the above work, I would also like to promote the following papers:

  • Johansson, F.; Brynielsson, J.; Horling, P.; Malm, M.; Martenson, C.; Truve, S.; Rosell, M.; , “Detecting Emergent Conflicts through Web Mining and Visualization,” Intelligence and Security Informatics Conference (EISIC), 2011 European , vol., no., pp.346-353, 12-14 Sept. 2011 [pdf]
  • Sitaram Asur and Bernardo A. Huberman. 2010. Predicting the Future with Social Media. In Proceedings of the 2010 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology – Volume 01 (WI-IAT ’10), Vol. 1. IEEE Computer Society, Washington, DC, USA, 492-499. DOI=10.1109/WI-IAT.2010.63 [pdf]
  • Ginsberg, Jeremy, Mohebbi, Matthew H., Patel, Rajan S., Brammer, Lynnette, Smolinski, Mark S., Brilliant, Larry. Detecting influenza epidemics using search engine query data. Nature. 2009/02/19/print. Macmillan Publishers Limited. [pdf]

Emergency Response Work as a Sociomaterial Practice

On thursday night at 21:20, my cellphone beeped and an SMS was received informing that the situation room was about to be manned at the fire and rescue services in Gothenburg. A sports arena was on fire and the response work was estimated to continue all night. I arrived to the main fire station at 21:50 and started a field-study and ended at 02:30. My focus this night was oriented on how the people working in the situation-room and the technology in use were intertwined. This particular incident provided very good insights on how people, procedures, roles and technology could be understood as entanglement in practice. In addition, it was also clear that technology is hardly used according the designers intention but reinvented and restructured in situated action. In return, the actions by the human actors were shaped by the material properties of the technological actors. Trying to separate human actors and technological actors seems difficult. I believe that we must view emergency response work as sociomaterial practice in order to move beyond the less meaningful discussion of technology vs method when it comes to exploring innovative conduct in future emergency and crisis response work.

I am looking forward to discuss entanglement in practice and sociomateriality with my fellow researchers during the ISCRAM2012 conference in Vancouver, 22-25 april.

This blogpost is heavily inspired by the following text:
Orlikowski, W. J. “The sociomaterialty of organizational life: Considering technology in management research.” Cambridge Journal of Economics 34 (2009): 125-141

Visualisering av samverkan blir allt viktigare

I samband med större händelser försöker de professionella aktörerna att som en del av hantera en rad omedelbara operativa problem också skapa bilder av vilka andra aktörer som är engagerade med den aktuella händelsen. Att hålla koll på samverkande organisationer är i den lilla händelsen relativt enkelt. Men när händelsen är mer komplex blir bilden av involverade organisationer svårare att hålla aktuell.

Kanske är det dags att skapa nya funktioner i existerande systemstöd som krishanteringsaktörerna förväntas att använda som också snabbt och kontinuerligt ger olika former av visualisering av samverkande samt mer indirekt involverade parter. Detta kan enklast ske genom två följande metoder:

1. Extrahera användningsstatistik i de vanligt förekommande loggningsystemen såsom WIS, LUPP-applikationen och Samverkanswebben
Skapa nätverksmodeller där relationer mellan aktörerna baseras utifrån hur information delas, vilken information som aktörerna läser.

2. Extrahera samtalsloggar från mobiltelefonerna som aktörerna använder samt kommunikationsloggar från Rakel. Skapa nätverksmodeller baserat på vilka roller och vilka organisationers som varit i kontakt med vem.

Bilden nedan visar en enkel visualisering baserat på en begränsad mängd användningstatistik från WIS i samband med Stormen Dagmar i december 2011. Vad som är intressant med bilden är hur informativa visualiseringar kan skapas med ganska enkla medel baserat på ytterst begränsad användningsstatistik.

Det är rimligt att anta att spännande och informativa visualiseringar för krishantering är möjliga på sikt när tillgången till större volymer och mer detaljerad användningsdata blir tillgängliga myndigheternas centrala systemstöd.
Svenska krishanteringsaktörer har mycket att vinna på att börja utnyttja den data som skapas som en bieffekt av att deras personal använder de centrala systemstöden. Syftet med sådana visualiseringar är att underlätta för krishanteringsaktörerna att tolka komplexa och svårgreppbara fenomen och händelseförlopp.

Särskilt intressant är att se hur loggar från sociala medier, mobiltrafikloggar, användningsloggar från WIS, loggar från biljettsystem i kollektivtrafik samt loggar från entrésystem i byggnader kan kombineras och ligga till grund för nyskapande visualiseringar.

Viktigt att nämna i detta sammanhang är att dessa loggar inte behöver kunna identifiera individer utan istället visa flödesvolymer där de enskilda data elementen alltså inte identifierar personer.

Users must have the app preinstalled in case of an emergency – or maybe not!

A well-known problem for everyone with a smart idea regarding how people could make use of innovative apps in case of an emergency or crisis, is the fact that people need to have these apps already installed ready to be used in critical situations. However, most people will never face an emergency or a crisis which means that most people will never bother to “pre-install” nice to have apps that most likely never will be used.

This means that app developers and solution providers must be come up with better solutions than just saying:

“People must understand that they have much to gain if they download our great app and have it ready at hand if they end up in a critical situation”.

Just recently, we have seen insightful approaches of solving this problem by designing apps that are made available when they are needed and only if they are needed. This means that app developers and solution providers have taken a new look at the well-known and standardized “sign-up -> receive user info -> get started and add details”.

Instead of using the above approach, innovations are emerging targeting what could be called a simplified user credentials approach. This means that apps could be designed using a range of new models concerning the role and need for user-id. When we design for “instant and short time use”, it opens up for radically different use cases and a potential for a much larger diffusion of apps in emergencies and crisis.

I hope that we at the Crisis Response Lab, in the near future, will be able to show a few examples of how these new approaches could be materialized.

Open positions at the Division of Interaction Design, Chalmers University of Technology

The Division of Interaction Design has three available positions at Chalmers University of Technology.

  • Associate Professor/Senior Lecturer in Graphical Interfaces (reference number 20120006).
  • Lecturer in Interaction Design (reference number20120005).
  • PhD student position in Computer Science: HCI and Visualization (reference number 20110141 ).

More information about the positions is available on: