Researchers based at the University of Cardiff are working on new artificial intelligence (AI) tools to help tackle hate crime as they surge to their highest levels ever recorded.
Over the past year, reported incidents of hate crime have risen by 17 per cent to 94,098 cases. Reported incidents have more than doubled since 2012-13 to the highest levels ever recorded. The majority of these (76 per cent) are related to race, followed by sexual orientation hate crimes (12 per cent), religious hate crimes (9 per cent), disability hate crimes (8 per cent) and transgender hate crimes (2 per cent).
A Home Office report on hate crime from 2017-18 stated that while increases in hate crime were largely driven by improvements in crime reporting, “there [have been] spikes in hate crime following certain events such as the EU Referendum and the terrorist attacks in 2017”.
Research carried out by the University of Cardiff’s ‘HateLab’ confirmed that spikes of online race and religious hate speech appeared in the hours and days after the Brexit vote, as well as incendiary events such as parliamentary votes, newspaper headlines and political speeches. Authorities are concerned that the UK’s scheduled departure from the EU in March 2019 could lead to a further spike in hate crime.
“Brexit has drawn sharp divisions in society, and the seemingly impossible promises made by leave-backing MPs have created a great sense of disillusionment in millions of citizens,” said Professor Matthew Williams, a Cardiff criminologist and principal investigator on the HateLab project.
“In 2019, Britain is likely to be in its most severe crisis in peacetime, and whatever the outcome, be it a second referendum, a soft-Brexit or a no-Brexit, there is concern that events will motivate more hate crime. As we saw following the 2016 vote, and to a more extreme extent following the 2017 terror attacks, surges in online hate speech coincided with significant increases in hate crimes offline.”
Now, HateLab researchers will be working to develop tools to help law enforcement and government predict where outbreaks of hate crime are most likely to erupt next.
The researchers – who use data science to understand hate crime and speech – will be monitoring online comments responding to political events and other newsworthy incidents, such as the terrorist attacks in Manchester and London, in order to develop an ‘Online Hate Speech Dashboard’.
“To date, there have been significant delays in getting information to staff following ‘trigger’ events,” said Professor Peter Burnap, computational lead on the project. “The dashboard will allow key personnel to gain aggregate insights into online reactions to events, such as our exit from the EU and terror attacks, in the so-called ‘golden hour’.”
The work is funded by the Economic and Social Research Council, and will be supported by the National Online Hate Crime Hub, which was launched in 2016 amid concerns following the Brexit referendum. The HateLab is also working in partnership with hate crime charities and police forces in Wales and Manchester.
Research from HateLab will be presented to Parliamentarians and senior civil servants at an All-Party Parliamentary Group meeting in early 2019.
In 2016, a group of Cardiff researchers led by Professor Williams developed a statistical tool to analyse social media and predict hate crime in Los Angeles. The tool aimed to help law enforcement prepare to take timely action against these crimes.
