ISCRAM 2017 Tutorials

Visual Analytics with Social Media for Emergency Response - Tutorial

Introduction to the Tutorial

A key success factor for Crisis Management is to achieve and maintain situation awareness, i.e. “accurate, complete and real-time information about an incident” [Winerman, 2009], to understand “the current local and global situation and how this may evolve over time” [Endsley, 1995]. Traditional approaches to situation awareness and crisis management tend to rely on official communication channels, which are generally slow in providing information, due to the need of releasing only information that has been verified and approved. However, information spreads very quickly by word of mouth, especially on social networks (e.g. Facebook and Twitter). This has become a major source of information for organizations, user groups and communities to understand important situations and events. Such data streams are particularly useful in the case of emergencies and crises - where highly critical and potentially life-saving decisions need to be taken in a short amount of time.

Visual Analytic systems can be used by professional users as well as citizens and communities to understand an emergency situation and its evolution. Such solutions provide an intuitive and immediate overview of a situation. Multiple visual representations support different information needs and preferences: a professional user may decide to look in depth at the timeline of an event whilst a citizen may just want to see the information that is relevant to their neighbourhood. Keeping with the theme of this year’s conference on ‘Agility is coming’, we will be investigating how real-time information management can be obtained through multiple and simultaneous visualisations.


[Endsley, 1995] Endsley, M. R. Toward a theory of situation awareness in dynamic systems: Situation awareness. Human factors 37, 1, 1995.

[Winerman, 2009] Winerman, L. 2009. Crisis communication. Nature 457(22):376-378.

Tutorial Topics

This Full Day tutorial is aimed at addressing three themes:

  1. Understand how to critically analyse visualisations.
  2. Gain technical knowledge on how to gather, analyse and visualise social media data.
  3. Group-based activities.

We will use freely available open source tools and libraries for data visualisation that can be easily used and customized by all types of users, even without specific programming experience. This hands-on tutorial will provide an overview of the various sources of information that can be accessed using such tools and a demonstration of how a visual analytics solution can be built using open source components will be provided. The final goal of the tutorial will be to provide every participant with means and know-how to collect, process, monitor and analyse social media.

During the tutorial, participants will be learning how large-scale data can be communicated to end-users. This will cover the process of analysing data, designing a solution to implementing it. Examples of good and bad visualisations and info graphics from real-world studies will be presented to the participants, with discussions on how to critically study visualisations and how they can be improved. This aspect is aimed at being more interactive, and will benefit the participants by learning about good and bad principles of design for visual analytic systems, as well how to inspect and examine visualisations.

A live exercise demonstrating the potential of a visual analytics solution will be carried out at the end of the session, where the different visual analytics solutions implemented by the participants will be tested against live data.


A basic knowledge of programming is needed for all participants. We welcome participants who can be PhD students or researchers interested in learning more about applying visualisations for exploring Linked Data and Social Media. Practitioners and industry professionals would also provide an excellent input in the course, to share their experiences and ideas. We welcome participants from different fields to help share a wide range of ideas and experiences, which is very helpful for the discussion and interactive sessions. Recruitment will be primarily via email lists and Social Media. The tutorial chairs are participants in ongoing or recently concluded large European projects on Citizen Observatories, Citizen Science, Big Data Analytics and Crowdsourcing. Additionally, the co-chairs are members of professional organisations, which provide a wider network for advertising the tutorial.

Requirements for Participation

Tutorial Structure

The tutorial will be run by the co-chairs. The start of the course will divide the participants into groups of two/three (based on programming expertise), depending on the total number of participants. Technical sessions will follow an interactive, hands-on approach, where one instructor will demonstrate and the other will provide support to the groups. Sessions will involve the instructor writing code on an online collaborative code-sharing platform. These will allow participants to login and access the code being written live. The code can then be copied/edited to be used in the empty/structured files available in the material provided. Non-technical sessions will be mostly interactive and discussion-based. Participants will be encouraged to share insights, examples and ideas from their relevant fields of research. These sessions will be led by one instructor, following a set of slides as an illustration, to seed further group discussions. Once participants have a working visual analytic system, a final group discussion will take place, as a breakout session to discuss among group members how to improve the system, and provide suggestions. Each group will then be invited to present their improvement, and finally the participants will choose a set of ideas and suggestions. These will be then shared with all participants, and the findings will be shared post-event, online in one of the websites hosted by the instructors and advertised on Social Media and any channels suggested by the course organisers.

We will provide a course template that includes:

  1. Set of slides that will be presented.
  2. Files (empty or following a standard structure) and folders to be copied to a particular local directory.
  3. Sample Social Media data, indexed in a Apache Solr datastore to be queried.
  4. Final solution for the Visual Analytic system, available from a download link shared with the participants during the tutorial.

Agenda Sunday May 21, 2017 — Room 0A45

Sunday 21 May 2017
08:00-09:00 Registration
09:00-09:15 Introductions of participants and organisers
09:15-09:30 Overview and introduction to
Visual Analytics and Interface Design
09:30-10:00 Technical Organisation
  • Organise working groups
    (2-3 participant per group, based on experience)
  • Downloading tutorial material, tools and obtaining API keys
    (if needed)
10:30-11:00 Coffee Break
11:00-11:30 How to access Social Media
  • Social Media Platforms and APIs
  • Gathering and Processing Social Media (practical)
  • Storing, Indexing and Retrieving data (practical)
12:30-14:00 Lunch
14:00-14:30 Visual Analytics Techniques for Social Media
  • Presentation of Visualisation techniques and technologies
  • Build a Visual Analytics system (practical)
15:30-16:00 Break
16:00-16:30 Interactive Exercise as breakout sessions
  • How can we improve? (Given the visual analytic system that was built, critically analyse the positive and negative aspects, ideas to new designs/visualisations)
  • Setup for ISCRAM 2017 (start an instance of the visual analytic system to monitor Social Media during ISCRAM 2017)

Tutorial Chair and Co-Chair

Dr. Suvodeep Mazumdar (
The University of Sheffield, UK

Dr. Suvodeep Mazumdar is a post-doctoral researcher in the OAK Group and holds a PhD in Computer Science from the University of Sheffield. His field of research concerns Human Computer Interaction, crowdsourcing [3] and Visual Analytics for organisational knowledge management and the social web [1,2]. He co-chaired tracks on Visual Analytics for Crisis Management and has run tutorials on Visual Analytics with Social Media for Crisis Management at ISCRAM 2013-2015

He was chair of the SMILE2013, SMILE2014 (Social Media and Linked Data for Emergency Response, co-located with ESWC conference) and VISUAL2014 workshops. He also organised a tutorial at ESWC2014, 2015 on Visual Analytics with Linked Open Data and Social Media.

[1] Suvodeep Mazumdar, Daniela Petrelli, Fabio Ciravegna. 2011. Exploring user and system requirements of Linked Data visualization through a visual dashboard approach.Semantic web: interoperability, usability, applicability.

[2] Suvodeep mazumdar, Daniela Petrelli, Khadija Elbedweihy, Vita Lanfranchi, Fabio Ciravegna. 2013. Affective graphs: The visual appeal of linked
data. Semantic Web–Interoperability, Usability, Applicability. IOS Press (to appear, 2014).

[3] Citizens Observatories for Effective Earth Observations: the WeSenseIt Approach. Suvodeep Mazumdar, Vita Lanfranchi, Neil Ireson, Stuart Wrigley, Clara Bagnasco, Uta Wehn, Rosalind McDonagh, Michele Ferri, Simon McCarthy, Hendrik Huwald, Fabio Ciravegna. Environmental SCIENTIST Journal. Volume 25, No. 2, 2016

Dr. Vitaveska Lanfranchi (
The University of Sheffield, UK

Dr. Vitaveska Lanfranchi is a Senior Lecture in Medical Computing at the University of Sheffield. Her research field concerns Human Computer Interaction with a focus on supporting, gathering and sharing of knowledge between individuals. She has chaired a track on Visual Analytics for Crisis Management and has run the tutorial on Visual Analytics with Social Media for Crisis Management at Iscram2013. She is an ISCRAM member and is also demos and tutorials co-chair of ISCRAM2015. She has published numerous papers in the area of the workshop, as well as several papers at ISCRAM editions since 2009.