Computational Fake News Analysis: A Practical Workshop

You are here

Computational Fake News Analysis:
A Practical Workshop

Date: 15-16 June 2017

Location: King’s College London, Strand campus, Franklin Wilkins Building FWB 1.16, WC2R 2LS LONDON

Free registration: please register at (password: FKNEWS1506):

https://www.eventbrite.co.uk/e/computational-fake-news-analysis-a-practical-workshop-tickets-34776359026

Aims and scope

The recent past has highlighted the influential role of social networks and other digital media in shaping public debate on current affairs and political issues. Disinformation and the alternative media ecosystem distort societal debates, increase polarisation, and threaten participatory democracy. For instance, the surprising success of Brexit and Trump’s election has been, at least partially, attributed to the unprecedented weave of false information that in both cases have polluted online debate before the vote.

Not only do fake news get significant attention and shares, but also alternative narratives often try to gain credibility through reusing content from mainstream media, often framed so as to undermine reader confidence in the latter.

The aim of this workshop is, firstly, to set out the state-of-the-art and challenges in fake news analysis, followed by hands-on practical sessions on relevant computational tools, including hyperlink crawling, network analysis, and rumour analysis.

Programme

15 June 2017

10:00 - 10:30 Workshop introduction and aims (K. Bontcheva & M. Cotè)

10:30 - 12:30  Panel Discussion on Fake News Analysis: State-of-the-art and Challenges  

Chair:

Tobias Blanke (King’s College London)

Martin Moore (Centre for the Study of Media, Communication and Power, KCL)

First Draft News representative (TBC)

Kalina Bontcheva (University of Sheffield)

  • Rumour analysis in the PHEME project

Mark Cotè (King's College London)

Mathieu Jacomy (Sciences Po, Paris)

  • Field guide to fake news

 12:30 - 13:30  Lunch and Networking

13:30 - 17:30 Investigating Fake News Site Networks

Learn how to use open-source hyperlink crawling and network analysis in order to investigate the connectivity profile of fake news websites. During the workshop, we will crawl the ‘fake news ecosystem’ as a multipartite domain inhabited by different types of websites (that may or may not be actually labelled as fake news) such as:

  • political satire
  • conspiracy theory websites
  • traditional news
  • extreme political websites
  • websites farming ad revenues
  • fake news trackers
  • actors who produce discuss about the fake news phenomenon

Participants will learn how harvest the network of hyperlinks connecting these websites and how to visualise and analyse such network.

16 June 2017

09:30 - 12:30 Analysing Online Rumours

Learn how to use open-source text and social media analysis tools, in order to detect automatically online rumours, analyse opinion polarisation around these, and determine their likely overall veracity. This session will present the latest tools developed in the PHEME project on computational rumour analysis.  

12:30 - 13:00 Closing remarks

Registration

Attendance is free, but please register at (password: FKNEWS1506):

https://www.eventbrite.co.uk/e/computational-fake-news-analysis-a-practical-workshop-tickets-34776359026

 

Tools (Please install in advance, if possible)

  • Gephi (network analysis)

http://hyphe.medialab.sciences-po.fr

  • GATE (rumour analysis; text and social media analytics)

https://gate.ac.uk/download/#latest