Second workshop on:

Exploitation of Social Media for Emergency Relief and Preparedness (SMERP)

23rd April, 2018

Co-located with:

The Web Conference 2018 (formerly WWW),
23rd - 27th April, 2018,
Lyon, France

HTML5 Icon Venue: Salle St Clair 1


  • SMERP 2018 workshop took place on 23rd April, 2018. Thanks to the all the keynote speakers and the participant speakers for making this event memorable and insightful. Thanks to the organizers -- Saptarshi, Debasis and Prof. Moens for conducting the sessions.
  • The program (schedule) is below. Please also find the tentative venue above.
  • Paper notifications sent. Five papers accepted - the list is here. Congratulations to the authors!!
  • We are delighted to announce that Prof. Prasenjit Mitra, Professor, The Pennsylvania State University, will deliver a keynote talk. The abstract of the talk is here.
  • We are glad to announce a keynote talk by Prof. Amit Sheth, Wright State University. The abstract of the talk is here.
  • Some clicks

    Prof. Prasenjit Mitra (Keynote 1)

    Prof. Prasenjit Mitra

    Prof. Amit Sheth (Keynote 2)

    Prof. Amit Sheth

    SMERP 2018: some of the attendees and organisers



    Session 1
    09:00 - 09:30 - Opening
    09.30 - 10.30 - Keynote Address by Prasenjit Mitra: Utility of Social Media Information in Response to Natural Disasters [abstract][slides]
    10:30 - 11:00 - Tea/Coffee Break
    Session 2
    11:00 - 11:20 - [Accepted Paper] Modeling evacuation behavior of NYC Twitter users during Hurricane Sandy. Dheeraj Kumar and Satish Ukkusuri [pdf]
    11:20 - 11:40 - [Accepted paper] Flood relevance estimation from visual and textual content in social media streams. Anastasia Moumtzidou, Stelios Andreadis, Ilias Gialampoukidis, Anastasios Karakostas, Stefanos Vrochidis and Yiannis Kompatsiaris [pdf]
    11:40 - 12:00 - [Accepted paper] Class Specific TF-IDF Boosting for Short-text Classification. Samujjwal Ghosh and Maunendra Sankar Desarkar [pdf]
    12:00 - 12:20 - Discussion: Introducing the questions for panel discussion
    12:20 - 13:30 - Lunch break
    Session 3
    13:30 - 14:30 - Keynote Address by Amit Sheth: Transforming Social Big Data into Timely Decisions and Actions for Crisis Mitigation and Coordination [abstract][slides]
    14:30 - 14:50 - [Accepted paper] Gold Standard Creation for Microblog Retrieval: Challenges of Completeness in IRMiDis 2017. Ribhav Soni and Sukomal Pal [pdf]
    14:50 - 15:10 - [Accepted paper] SAVITR: A System for Real-time Location Extraction from Microblogs during Emergencies. Ritam Dutt, Kaustubh Hiware, Avijit Ghosh and Rameshwar Bhaskaran [pdf]
    15:10 - 15:40 - Tea/Coffee Break
    Session 4
    15:40 - 16:40 - Panel discussion: Answers to the questions
    16:40 - 17:00 - Open discussion and closing

    Call for Participation

    The ever-increasing amounts of user-generated contents on online social media (OSM) platforms such as Twitter, Facebook, WhatsApp, etc. have become important sources of real-time information during emergency events (e.g. natural disasters like earthquakes, cyclones, floods, fire, epidemics or man-made disasters like terror attacks, riots). During such an event, various information is posted on OSM, which can contribute significantly to relief operations. Additionally, crowdsourced content from OSM can also be utilised for emergency preparedness, such as for identifying disaster-prone regions and infrastructures, developing early warning systems, developing emergency-resilient communities, and so on.

    Given the huge volume and the rapid rates at which content is posted on OSM, automated techniques and information systems need to be developed for extracting, summarizing and presenting the critical information in a useful way. The proposed workshop aims to provide a forum for researchers working on related fields to present their insights. The workshop aims to bring together researchers from diverse fields -- Information Retrieval, Data Mining and Machine Learning, Natural Language Processing, Social Network Analysis, Computational Social Science, Human Computer Interaction -- who can potentially contribute to utilising social media for emergency relief and preparedness. One specific objective of SMERP 2018 will be to promote multi-modal and multi-view information retrieval, i.e., developing methods for aggregating information from multiple online and offline data sources (including text, images, and video).

    The secong edition of this workshop will consist of two peer-reviewed tracks: (1) a general track, and (2) a data-focused track, briefly described as below:

    General Track

    This track of the workshop aims to provide a research platform to explore the role of social media in emergency relief and management and requests for original research contributions related to the theme Exploitation of Social Media for Emergency Relief and Preparedness. The detailed aims and scope of this track are provided here.

    Data-focused Track

    Here we provide a dataset comprising of (i) tweets, (ii) images, and (iii) news reports, all published during a particular emergency event - the 2015 Nepal earthquake. We invite interested participants to develop methods to solve some practical challenges over the dataset.

    More details are given here.    

    General Track (details)

    Aims and Scope

    The general track will solicit contributions related to the theme, which includes (but is not limited to):

    • Multi-modal and multi-view Information Retrieval - developing methods for aggregating information from multiple online and offline data sources (including text, images, and video)
    • Addressing the code-mixed and noisy, informal vocabulary of OSM content
    • Transfer learning - applying models trained on prior emergency event(s) on a future event
    • Detection of events and emerging themes
    • Real-time management and summarization of dynamic content streams
    • Detection of rumours and false news, verification of news and identification of trustworthy sources
    • Geo-tagging and geo-localisation of content and sources
    • Social network models for information diffusion in emergency situations
    • Identifying disaster-prone or accident-prone regions and infrastructures
    • Designing crowdsourcing systems for emergency preparedness and post-disaster relief
    • Building knowledge bases for automatically mining social media posts during emergency

    We encourage submission of novel work-in-progress papers that show promising directions, and papers that demonstrate systems that can be practically useful during emergency events.

    Submission details

    Important dates

    Data-focused Track (details)

    In the data-focused track, we provide a dataset comprising of the following data related to the 2015 Nepal earthquake:

    • A set of microblogs (tweets) posted during the event
    • A set of images posted along with the tweets
    • A set of news articles about the event, posted on various mainstream news media sites

    We invite interested participants to use this dataset to develop methods for the following two tasks:

    Task 1: Identifying information about infrastructure damage (the information might be in the text of a tweet or a news media article, or in the image associated with a tweet - multi-modal IR algorithms to be developed)
    Task 2: Verification of information posted on social media, using news articles (e.g. for a given social media post, develop methods to identify news articles which support or oppose the information in the post)

    Additionally, participants are welcome to use the released dataset for addressing some other problem (other than the two tasks specified above) relevant to the workshop theme.

    Those who submit to the data-focused track must use the released dataset (they will be free to use any other data as well).

    Submissions to the Data-focused track must make all their code and results available, e.g., on Github. The submitted paper must contain a link to the Github repository, so that reviewers can check their code and results.

    Getting the dataset

    To get the dataset, interested participants should fill up the data release form and email a scanned copy of the filled form to one of the organisers Kripabandhu Ghosh (kripa [dot] ghosh [at] gmail [dot] com) or Saptarshi Ghosh (saptarshi [dot] ghosh [at] gmail [dot] com). The organizers will send the dataset by replying to the mail.

    Submission details

    Important dates

    Organizing Committee

    Marie-Francine Moens, KU Leuven, Belgium
    Gareth Jones, Dublin City University, Ireland
    Saptarshi Ghosh, IIT Kharagpur, India
    Debasis Ganguly, IBM Research Labs, Dublin, Ireland
    Tanmoy Chakraborty, IIIT Delhi, India
    Kripabandhu Ghosh, IIT Kanpur, India

    Technical Program Committee

    Arnab Bhattacharya, IIT Kanpur, India
    Charles Jochim, IBM Research, Dublin, Ireland
    Dongxiao He, Tianjin University
    Douglas W. Oard, University of Maryland, College Park, USA
    Girish Keshav Palshikar, Tata Consultancy Services, India
    Jaegul Chao, Korea University, Korea
    Manish Gupta, Microsoft Research, India
    Monojit Choudhury, Microsoft Research, India
    Muhammad Imran, Qatar Computing Research Institute, Qatar
    Noseong Park, University of North Carolina, Charlotte, USA
    Parantapa Bhattacharya, Virginia Tech, USA
    Parantapa Goswami, Viseo Technologies, Grenoble, France
    Parth Gupta, Amazon, India
    Richard McCreadie, University of Glasgow, UK
    Rishiraj Saharoy, Max Planck Institute for Informatics, Germany
    Srijan Kumar, Stanford University, USA
    Tinne Tuytelaars, KU Leuven, Belgium
    Yufang Hou, IBM Research, Dublin, Ireland


    All submissions must be written in English following the ACM author guidelines.
    Full papers must not exceed 5 pages, and short papers must not exceed 2 pages, including all diagrams, references, and appendices. Submitted papers should include the names and affiliations of all authors.

    We solicit original contributions relevant to the workshop theme, that have not been published earlier or are not under submission at any other venue. Submitted papers will be peer-reviewed, and the accepted papers will be included in the conference companion proceedings.

    At least one author of each accepted paper needs to register for the conference and present the paper at the workshop.

    The papers for both the tracks should be submitted via Easychair. While submitting your paper through this Easychair link, please make sure to select "Exploitation of Social Media for Emergency Relief and Preparedness Workshop" as the track to which you are submitting.

    The data request related correspondence for the Data-focused Track will be done via smerp2018 [at] gmail [dot] com.

    For more details please see Call for Participation.

    Important Dates

    January 22, 2018 Paper submission due (extended)
    February 15, 2018 Notification of acceptance
    February 28, 2018 Camera-ready submission deadline
    April 23, 2018 Workshop

    All submission deadlines are at midnight Anywhere-On-Earth.


    At least one author of each accepted paper must register for The Web Conference 2018, and present the paper at the workshop.

    Past Events

    The 1st SMERP workshop was co-located with the European Conference on Information Retrieval, April 2017, at Aberdeen, Scotland.

    SMERP 2017 proceedings are available on CEUR here.


    For queries about the data challenge, you can mail smerp2018 [at] gmail [dot] com.

    For any other queries about the workshop, you can mail the following organizers:
    Saptarshi Ghosh: saptarshi [dot] ghosh [at] gmail [dot] com
    Kripabandhu Ghosh: kripa [dot] ghosh [at] gmail [dot] com