May 4, 2017

Let’s Talk about Trolls: Vuukle + Google to combat Trolls with AI Technology

Let's Talk about Trolls Vuukle Google Ai Fight Online Hate

Hindustan Times recently started a series called- Let’s talk about Trolls. This gave a motion to the dialogue and discussion on the much debated topic. Its necessary to understand that trolls are not harmless.

Many come to defense, whenever the topic of trolls comes up. Waving the banner of free speech and much-needed humor. Adding humor to a mundane debate is fresh. It is also one of the add-ons of using social media. But the same argument when used for the those using hate-speeches and abuses, is greatly foolish and ignorant.

Why Talk About Trolls?

Because it makes a difference.

Trolls – those vicious individuals who hijack any conversation by sowing discord, starting arguments, upsetting people, posting inflammatory comments, hurling abuses. Need to be tackled.

Someone like Sonam Kapoor gets trolled for iterating using national anthem, that the nation belongs to every religion. Gurmehar Kaur receives rape threats for voicing her protest. Newborn baby gets trolled for having a name not agreeable for some, what is his fault?

Typical Internet trolls

The Typical Internet Troll. Source:

Disagreeing with someone is one thing but using abuse, threats and hate to point it out, is the worst way. Fair criticism is part and parcel of life in public domain, but no one deserves abuses, hate and threats.

Everyone who has voiced a strong opinion on social media has got trolled, and those who haven’t, will. There’s no getting away from people who do not believe in using facts to win arguments but resort to abuse, insults and threats to hit out at people they disagree with.Dr. Samir Parikh

It is important to discuss the aspects and find solution to it.

Discussing Solutions:

How do we arrive at a suitable solution, without curbing free speech? How can a media technology company, the legal system and a speaker online limit hate speech and abuse? These are pertinent questions which need answers.

Facebook, Twitter, YouTube and Microsoft have signed up to the European Commission’s code of conduct. It includes a commitment to remove illegal hate speech on most valid notifications in less than 24 hours.

Thus, teams in Europe are better staffed. But, the same doesn’t happen in India. Similarly, different national laws prompt different responses in different countries.

Thus, building up Better laws is one of the solutions. The other major aspect of going about trolls is technology. Click To Tweet

Technology vs Trolls

The technological options used until now include- blacklisting keywords, reporting, flagging, removing undignified content and banning culprit users. This is all handled and taken care of by behind the scene heroes of the internet- Moderators.

Moderators are a human army using whatever tech is available to keep the internet as safe and sound as possible.They go about trolls, weed out hate speech, abuses, illegal content and the likes.

But with ever-growing number of users, humongous amount of data and growing popularity of digital medium, the moderators were never going to be enough. They need serious help.

The rise of Artificial Intelligence. Technology vsTrolls

The age of Artificial Intelligence. Technology vs Trolls Source:

It is here the role of algorithms and artificial intelligence comes into play. Google has developed an amazing artificial intelligence called Perspective.

Vuukle + Google : Teaming Up

Recently, Vuukle partnered with Google to integrate the Perspective API into its commenting system. Vuukle is the fastest growing engagement and commenting platform. This move comes as a great step for its anti-troll stance.

Vuukle is world’s first commenting system to integrate real-time author feedback with Google technology to automatically detect hate speechC J Adams, Product Manager, Google

The AI tool grades the toxicity in the comments on a scale of 1-100. The editor/moderator can set the toxicity level as required, say 65 . In that case, no comment with toxicity above 65 will get published. It will either be directed to editors or be eliminated outright.

Comment below and check out the system work its magic.

The specialty of the technology is that it needs no human effort. All of it happens real-time when the comment is still being typed. This provides better efficiency in terms of both speed and accuracy. No need to ban accounts as well. The stuff just won’t get posted if it is toxic. As a result, users have to adapt and post acceptable stuff.

Vuukle has made the Google technology available to more than 1000 digital publishers worldwide.

No usual techniques such as account changing, fake anonymous accounts will make any difference. It is a long way to go.. but the real battle has just started. The Vuukle-Google partnership right now is the quintessential Thor's hammer! #WarAgainstTrolls Click To Tweet

Santoshkumar Pandey

Engineer by qualification, content & marketing expert by profession. Meditate, blibliophile & a creative geek.