AI to help spot online bullies before they do their damage – Yahoo and IBM collaborate to tackle the scourge of online bullying head on
(Image credit: Shutterstock)
If you’ve used Twitter, it is clear that there is no dearth of anonymous users whose sole aim appears to be trolling people, even bullying users to the extent of causing serious mental and psychological effects. Twitter has been working with the likes of IBM to counter the proliferation of such instances using the latter’s AI prowess — their IBM Watson supercomputing platform.
With the intent of making the entire Twitter experience safer, the micro-blogging service recently announced in a blog post they they were working on ways to give more control to users over what they see on their timeline, as well as improving methods victims can utilize to report abusive tweets.
“We’re taking steps to identify people who have been permanently suspended and stop them from creating new accounts,” explained the post. “This focuses more effectively on some of the most prevalent and damaging forms of behavior, particularly accounts that are created only to abuse and harass others.”
Addressing the subject of what’s next in the fight against abusive Twitter users, the company explained that they are working with IBM’s Watson team, aiming to tap into the supercomputing platform’s ability to detect nuances in language, sentence construction, word choice and eventually infer intention. The company aims to incorporate such AI-backed approaches into its ongoing effort to quell the number of instances of abusive users and posts. As Twitter’s vice president of data strategy Chris Moody stated at a conference earlier this week, “What we want to do is be able to identify abuse patterns early and stop this behavior before it starts.”
As part of their Cognitive Build competition, IBM partners also came forth with ideas such as implementing an automated chat bot that could be deployed in cases of bullying, and to even counsel users who have been mentally affected by online bullying.
A recent IBM blog post explained, “We are starting with governments, non-for-profit organizations, and universities to create a counseling app. We are looking at how we can train Watson in the best counseling techniques, to have consistency and the most effective next steps. Data will be collected at every single stage, and Watson will continue to learn the most appropriate statements, approaches and suggestions to offer up. It’s in the very early stages.”
This isn’t the first time Twitter has collaborated with IBM to mine key information for improving the quality of their service. As early as 2015, Twitter worked with the Watson team to help extract relevant business insights from the feed using the AI’s analytic capabilities. Using IBM’s AI-powered service, it is possible to delve into Twitter conversations and dredge up key insights about users including the geographical location where a particular Twitter trend originated, querying the dataset with questions like “show me male to female ratio of users” or “How often does <keyword> appear in the conversation”, as well as viewing additional query filters for viewing Twitter data through.