Marie Truelove

[vc_row][vc_column width=”1/3″ css=”.vc_custom_1453342300079{padding-top: 15px !important;}”][vc_column_text]Marie Truelove[/vc_column_text][/vc_column][vc_column width=”2/3″][vc_column_text]

Research@Locate Presentation:
Introducing a Framework for Automatically Differentiating Witness Accounts of Events from Social Media

Marie Truelove, PhD Candidate at The University of Melbourne


[/vc_column_text][/vc_column][/vc_row][vc_row css=”.vc_custom_1453341977865{border-top-width: 1px !important;border-top-color: #28b7ff !important;border-top-style: solid !important;}”][vc_column][vc_column_text]

Things You Will Learn

  1. Whether an event is a natural disaster or a scheduled television spectacle, identifying the fraction of micro-bloggers posting Witness Accounts has applications in many domains from emergency response to marketing
  2. The number of Witness Accounts identified can be significantly increased from those with geotags alone, by additionally considering text and image content within micro-blogs
  3. These additional contents can also be considered as evidence to test whether they corroborate the categorisation of a Witness Account, or in fact raise doubt because they are in conflict
  4. Progress towards automation is summarised in this paper, including supervised machine learning techniques for the categorisation of text and image content, and Dempster-Shafer Theory of Evidence modelling for testing corroboration or conflict.

Target Audience

Researchers from a range of disciplines with interest in crowdsourcing from social networks, geographic information retrieval, machine learning and natural language processing
Industry professionals with interest in crowdsourcing from social networks, active in a range of industries including emergency response, journalism, and marketing.

Presentation Overview

Identifying Witnesses of events from social media is an opportunity to crowdsource real-time information to enhance numerous applications including emergency response in a crisis, filtering sources for journalism, and enhancing marketing products. Using a sporting event broadcast live to a proportionally much larger audience, this research demonstrates a significant increase in the number of Witnesses identified posting from the event venue, in comparison to the number identified from geotags alone. This is achieved by considering the text and image content of micro-blogs as additional evidence. This paper also reports progress towards the automatic categorisation of the additional text and image evidence, and modelling and testing this evidence for corroboration or conflict, using Dempster-Shafter Theory of Evidence.


Marie Truelove is currently pursuing a PhD on characterising and distinguishing Witnesses of events from social networks such as Twitter, enhancing her expertise of the interactions between people, spatial science and technology. Marie has significant industry experience specialising in product managing emerging spatial technologies in start-up environments including previous roles at Location-based Services companies.