This is a guest post
The UK’s most active internet trolls live in two small suburbs of Kingston upon Hull according to new research by a software company that has developed an online protection program for children.
Software development company KevTech Apps has discovered that Marfleet, a suburb of Kingston upon Hull and Greatfield – a housing estate in east Hull – have the most active internet trolls per capita than anywhere else in the UK.
Both were identified as the worst postcodes in the UK for internet trolls – those who post threatening, offensive or derogatory comments on social media. They were closely followed by Severn Beach in Gloustershire and Westbury in Wiltshire.
Marfleet has a population of about 13,600 and its residents are almost all UK born. Its top occupations are listed as elementary. It shares three traits with a small town in Australia – Laidley in Queensland – which was identified by KevTech as Australia’s worst for internet trolls.
Both the UK and Australian towns had a small population, their residents were overwhelmingly born locally and had lower than average standards of education.
Of the top ten locations in the UK for the most active internet trolls, first five were in England followed by Crieff, a small market town north west of Edinburgh. Strabane in Northern Ireland ranked at number ten.
The software firm measured internet troll activity by using Twitter’s longitude and latitude references.
The developers collected data from Twitter’s public users and used Twitter’s geo location data to pinpoint those who posted material that might be harmful to others.
“We collected about 1.5 million tweets in total. We then used our contextual detection ability to identify tweets which contain threatening, derogatory or offensive content,” said Anil Chatterjee, CEO of the ground breaking Australian social media safety software, SafeKidsPro. The program alerts parents and carers if children are the targets of cyber bullying, online predators or other threats.
As children embark on a new school year and as a result of the inevitable peer group pressure and the longing to belong, many of them will join social networks for the first time. For a great many of these children (52% according to no bullying.com) it will also be the first time they become exposed to internet trolls, cyberbullies and other faceless people who would do them harm. SafeKidsPro has been developed to enable parents to watch out for their kids on social media as they would in the physical world.
SafeKidsPro’s detection ability was born out of a research project conducted with the University of Queensland to see if machine learning techniques could analyse messages in context and better detect cyberbullying than the existing commercial solutions that rely on simply matching of key words and phrases.
To do this, KevTech took 5 million messages from social media and gave a random sample of 100,000 to members of the general public to analyse in context and define as being “good”, or of no interest, or “bad”, and containing threatening, sexual, offensive, derogatory or swearing content, and importantly, intent.
Using the results of this analysis, algorithms were developed and formed the basis of SafeKidsPros’s detection ability.
The University of Queensland tested SafeKidsPro’s systems against the number one social media safety product, Net Nanny, and concluded that SafeKidsPro was 20-30% better at detecting “bad content.
We also harnesses unique artificial intelligence technology which emulates brain function to determine how someone may feel after receiving repeated “bad” messages.
It can also determine the potential intent of the person sending those messages. In this way it can alert parents to potential cyber bullying whether their child is the victim or the perpetrator.
Using this same method, we can also alert parents to potential predatory interactions such as grooming, stalking and sexploitation. This gives parents the opportunity to intervene early and minimise any potential damage caused.
We offer many other safety features like allowing you to see who your kids is “friends” with on Facebook and “following” or “followed” by on Instagram and Twitter. It allows parents to view images shared to check their suitability and it will alert them if their child shares pre-defined private information, or makes their location and more.
“At KevTech we are technologists but first and foremost we are parents”, said Co-founder and CEO, Anil Chatterjee. “We’ve become increasingly alarmed and saddened by the negative impact of bad social media behaviours, conscious of the lives it can damage or destroy in the short and long-term. SafeKidsPro is about keeping your kids and other kids safe where you cannot physically be. It’s about managing their reputations when they don’t appreciate the implications of their actions. It’s about responsible parenting in the digital age. All this for less than 50 pence per week and we even offer parents a free trial. We believe it is sound value. ”
How about you, have you come across internet trolls? I know I have and it’s frightening but technology like this is so valuable. Thanks to Anil for guest posting today, you can contact him at: