2nd workshop “Social Media for Disaster Risk Management (SMDRM): Researchers meet Practitioners”

Social media has been described as a form of distributed knowledge, a mechanism for understanding a situation using information spread across many minds. The interactions among people in social media are a form of collective intelligence as they allow people to make sense of a developing event collectively. 
Social media users can contribute to the creation of a "sensor" for citizen-generated data that modelling or monitoring systems can assimilate during a crisis. At the JRC, we think that social media constitutes a growing data source to help improve response in the early hours and days of a disaster, when gaining situational awareness is critical and time sensitive. However, social media platforms may not provide the functionality of summarizing the information that is useful for crisis responders.
During the 1st workshop on SMDRM, held in November 2020, practitioners, although widely recognizing the potential value of social media for accessing timely information, outlined some critical challenges for improving its adoption during crises. These challenges include validating and integrating near real-time information generated on social media with authoritative information and more traditional information systems; and preventing negative impacts from misinformation and disinformation. 
All participants expressed their desire to continue the discussion towards identifying new directions for research and development of systems that can better serve the information needs of emergency managers. We (JRC) have listened, and we assembled a task force with the aim to identify, understand, and address the challenges of improving the adoption of social media data for disaster management by taking a collaborative approach to co-design solutions.
The number of participants increased from 70 (1st workshop) to 105 during the most visited panel of the second event. During the workshop, practitioners presented operational tools deployed during emergencies coordination activities (impact analysis for GDACS) during Haiti’s 2021 earthquake and innovative tools for automated landslide impact detection). 
They also discussed what were the main barriers to the effectiveness of social media during critical times: whether it was fast or reliable enough. Across the sessions, practitioners shared their insights, experiences, and views on social media use in recent incidents. They described how they used social media data to gather situational awareness and to establish a communication channel with communities impacted by events such as the Beirut blast in 2020 or wildfires in Portugal. 
Emergency managers described how they managed to tackle the issue of understanding unstructured data by interacting with tech-savvy volunteers (VOST in the EU and CERT in the USA). And researchers presented new developments for improving the automated classification of signals from social media and the summarization of information. It is remarkable how much research work is also trying to extend the application of such tools to multiple languages, overcoming the critical issue of scalability for their operational use.
Interesting collaborations between researchers and practitioners were also presented and showed how both communities can benefit from them. Some practitioners are involved in studies to help define training data for researchers' models or evaluating them. Researchers also gather information about disasters with the help of practitioners to build reference datasets and establish platforms for favoring new developments (TREC Incident Streams)

Conclusions
Participants recognised how social media are not a silver bullet, they should be merged or complemented with other signals. However, when social media data are structured, filtered, and trusted, they can provide information that crisis managers need the most. Ground truth information from traditional systems can be late and earth-observation tools cannot catch the dynamic of an event.
As of today, fully automated systems for the collection of data are available and mature, and AI-supported systems for filtering the information are well tested. To assure the achievement of the last mile for having social media in the crisis room as any other physical sensor, practitioners need to be involved early on and throughout the research process. As with any new tool available to users, also the social media sensor needs to be introduced to crisis managers ready-to-use, and at the same time crisis managers should acquire the skills to properly use it.
We look forward to continuing and strengthening the relationship between research and practice in the coming years. The task force besides the workshop has launched a survey targeted at practitioners with the goal of understanding their needs and their use of social media to identify the usefulness but also the biases of such data. They would like to explore the possibility of a real-time exchange of expertise and data between researchers and practitioners using a common data model. To access to the workshop presentations or to collaborate for further activities, you can contact the team via: https://drmkc.jrc.ec.europa.eu/initiatives-services/social-media-driven-disaster-risk-management

Link:  https://drmkc.jrc.ec.europa.eu/initiatives-services/social-media-driven-disaster-risk-management