PHEME – Computing Veracity Across Media, Languages and Social Networks

This webpage hosts a series of live graphics showing the number of stories appearing in online news media sources, over the last two months, relevant to common mental health concerns. It aims to provide mental health professionals, service users and the general public with accurate information about recent online news media coverage.

The page has been developed as part of PHEME – a major international research project to understand how facts, half-truths and deception spread on the internet and how to counter this with better quality information.  We need to identify stigma and misinformation in online news media in order to counter it and this webpage is part of that goal.

The consortium

The PHEME project is a multinational, research consortium of nine collaborating centres: five universities– University of SheffieldMODUL University ViennaKing’s College London, Universitaet des Saarlandes and University of Warwick and four companies – ATOS Spain SAUshahidi Inc, Swissinfo.ch and Ontotext AD.

Veracity (truthfulness) – a Big Data challenge

Facts, half-truths, lies and deception all occur in online media. When such information spreads rapidly, it can have direct and extensive public consequences. To understand how information spreads online, large amounts of online content need to be analysed quickly. We coined the term phemes to describe memes (ideas, behaviours and styles that spread within a culture) with a measure of truthfulness and deception; also a nod to Pheme, the Greek goddess of fame and rumours.

Veracity intelligence system

With funding from the European Commission, we have been developing a system to determine the truthfulness of information as it travels across media, social networks and languages. Specifically, we have been concentrating on four types of rumours: speculation, controversy, misinformation and disinformation. Our aim is to combine big data analysis with advanced linguistic and visual methods to produce a ’rumour intelligence’ tool for direct application in medical information and digital journalism systems. The project runs until the end of December 2016.

The tool we are developing in the PHEME project is being tested in two domains: digital journalism with our colleagues at Swissinfo.ch (to find out more please listen to our podcast and watch our video) and mental healthcare here at the NIHR Maudsley Biomedical Research Centre (BRC). 

Public health professionals increasingly use information from online media to study health-related attitudes and behaviours such as tobacco smoking, HIV prevention and medication effects. However, truthfulness, rumours and spam are major challenges.

At the NIHR Maudsley Biomedical Research Centre, PHEME will investigate online media (including news and social media), providing information and rumour intelligence for use by service users and healthcare and public health professionals.

The ability to spot media trends and rumours automatically could provide early alerts of issues that are likely to be raised by service users or the general public. Clinical advice and practice could then be revised earlier, resulting in more successfully tailored and effective consultations. The detection and identification of healthcare rumours will also help improve educational material and awareness campaigns as well as inform appropriate training for healthcare staff.  

To find out more, please listen to our podcast and watch our videos.

To see the relative numbers of media stories on each topic, please look at the ‘Topic Graph’ (the doughnut–shaped graph, top left below). This shows the proportion of stories that are on each topic published over the last two months.

Each of the other graphs shows online news media trends for each mental health concern in the last two months and up to the current date. Because this relies on keywords, it is a conservative estimate and the true numbers are likely to be higher. Clicking on individual graphs highlights the relevant keyword searches and associated trends for a given point in time.

We expect stories to increase and decrease over time driven by events such as mental health awareness days or changes in legislation, which may influence the type and volume of online news media associated with a specific mental health issue.

The data that each graph represents are drawn from a larger interface, developed by our colleagues at MODUL University Vienna, which identifies, tracks and analyses media topics in real time. For more information and to view news trends and visualisations, please visit our online dashboard. The interactive PHEME dashboard provides access to the latest trends in news and social media coverage about mental health. Its real-time visual analytics capabilities allow exploring the gathered content along geospatial, semantic and temporal dimensions. An overview providing detailed description of these capabilities, both in terms of data services and visual tools and instructions on how to navigate the dashboard can be found at http://www.weblyzard.com/interface.

For more information on the PHEME project, please contact contact@pheme.eu.

If you are having viewing difficulties or other technical issues, please contact feedback@weblyzard.com

If you are worried about your mental health, or that of somebody you know, and not already receiving support, you should contact your GP (family doctor) to discuss your concerns. Your GP will have experience with most mental health topics that are referred to on this webpage and may be able to help you directly or put you in contact with more specialist mental health services.

If you urgently need to speak to someone, the Samaritans are available 24 hours a day, seven days a week on 116 123.

Mind also provides information concerning mental health issues and runs several helplines. Please visit http://www.mind.org.uk/information-support/helplines for details.

 

Topic coverage

Dementia

Self Harm

Autism Spectrum Disorders

 

Schizophrenia/Psychotic Disorder

Attention Deficit Disorder

Bipolar Affective Disorder

Depressive Disorder

 

Alzheimer’s Disease

Anxiety Disorder

Obsessive-compulsive Disorder

 The data provided in this section of the website are powered by webLyzard technology and the NIHR Maudsley Biomedical Research Centre is not responsible for accuracy of the data or technical issues you may encounter.