Email
Password
Remember meForgot password?
    Log in with Twitter

article imageNew AI aims to remove racism online

By Tim Sandle     Jul 31, 2020 in Internet
A company is making it their mission to remove racist words from the web. The firm UserWay has announced the launch of an artificial intelligence tool called Powered Content Moderator Tool, designed to remove inflammatory speech from the Internet.
The tool works by flagging discriminatory, biased, and racially-charged language content for site administrations to review and approve. It is an enormous task for most site owners to keep track of the latest consensus around culturally sensitive terms, and UserWay makes this previously almost impossible task simple, centralized and scalable.
Prior to launching this new tool, UserWay reviewed more than 500,000 websites for discriminatory content. This included most widely-used and popular websites. Among the findings was the fact that 22 percent of websites reviewed contained some form of biased, racially charged or offensive language. Of those:
52 percent of websites were found to contain instances of racial bias.
24 percent of websites had instances of gender bias.
12 percent of websites were found to contain instances of age bias.
There were websites with racial slurs (at 5 percent)
3 percent of websites were found to have a form of disability bias.
In terms of what triggered the suspect content, the most frequently flagged words by this tool for racial bias were:
“Blackmail”
“Whitelist”
“Black sheep”
“Blacklist”
“Black mark”
Such terms are particularly important in the context of 2020's Black Lives Matter social movement.
Those terms most often flagged for gender bias included words like: “chairman,” “fireman,” “mankind,” “forefather,” and “man-made.”
Also added to the list of suspect words are derogatory terms associated with COVID-19, such as “Covidiot,” “China virus,” “Kung-flu,” and others.
Some websites are adopting better practices. This is particularly so with marketers and technology companies who have taken strides to update their content, copy and computer code to terms that are racially and gender-neutral.
More about Artificial intelligence, Racism, inflammatory, Hate speech
 
Latest News
Top News