Lansons

Lansons Conversations

Managing reputation by detecting fake information on social media

In the USA, Dow Jones plunged 140 points after a rumour spread on Twitter from Associated Press’ Twitter account. The estimated temporary loss of market cap in the S&P 500 totalled $136.5 billion.

 

When hackers took over Associated Press’ Twitter profile in 2013

When hackers took over Associated Press’ Twitter profile in 2013

This isn’t the only example of misinformation spread via social media that has had cataclysmic real-world consequences; the England Riots spread violence, false information around the Ebola outbreak caused increased deaths (salt water does not prevent or cure Ebola), and the Boston marathon blasts identified the wrong suspects.

The spread of information via social media can have real-world consequences and the notion of ‘influence’ may spell the downfall or uprising of an organisation. This is why appropriately detecting, tracking, and engaging with false information on social media is critical for managing reputation online.

In a new paper published by Aditi Gupta from the Indraprastha Institute of Information Technology Delhi, social media reputation management techniques are explored in detail. Over 25 global events are analysed between 2011 – 2014 that involved the spread of fake images, rumours, and untrustworthy content.

“Online social media has the capability of playing the role of, either a life saver or that of a daemon during the times of crisis. In this research work, we highlighted one of the malicious intended usage of Twitter during a real-world event, i.e. spreading fake images. We analyzed the activity on the online social networking website Twitter, during Hurricane Sandy (2012) that spread fake images. We identified 10,350 unique tweets containing fake images that were circulated on Twitter, during Hurricane Sandy.

We performed a characterization analysis, to understand the temporal, social reputation and influence patterns of the spread of these fake images. We found that 86% tweets spreading the fake images were retweets, hence very few were original tweets by users. Also, our results showed that top 30 users (0.3% of the users) resulted in 90% of retweets of the fake image.”

 

Rumour tweets posted during the England riots of 2011

Rumour tweets posted during the England riots of 2011


The research concluded that after analysing some of the top disasters over the last four years, only automated techniques were able to successfully identify credible updates and categorise. Of course, this was the only option as manually sifting through 5.6 TeraBytes of tweets would take a lifetime! At Lansons we use our own reputation management tools.

When managing online reputation you should watch out for:
• The creation of fake social media profiles that are designed to look real but instead spread fake information;
• Fake content, then engage, before it’s spread widely on social media;
• Online communities that build around disasters that could be considered the ‘core’ group that drive wider conversations.

The research is a valuable contribution to managing online reputation, it allows practitioners (such as myself) to refine and improve our techniques. Whilst the focus of the paper was on Twitter, we know reputation management applies to all social media – even more prominently in Google Search. Make sure you have the right procedures in place.