Meet the Internet's new cops

Jun 03, 2018, 07:05 IST | Anju Maskeri

A content moderator of nine years tells us what it takes to make the Web a safe place

(Extreme left) Rajeev Srivastava with his team. Pic/Ajeesh Rawther
(Extreme left) Rajeev Srivastava with his team. Pic/Ajeesh Rawther

On an average, 29-year-old content moderator Rajeev Srivastava at Bengaluru's Foiwe Info Global Solutions, reviews approximately 16,000 items a day including text, images and videos. During the nine-hour rotational shift — this week it starts at 1 pm — Srivastava's task is to discern whether the content, most of which is user-generated, is okay to be published online.

It has been his job for the last nine years after a short, two-month stint as content writer in Noida. But monotony is the least of his concerns. "Frankly, when I was hired, I wondered whether I would be able to handle the onslaught of content, some of which can be disturbing, on a regular basis. People are hungry for attention, so they might upload videos of violence or unspeakable things," says Srivastava, a B-Tech graduate from Varanasi. But the two-month-long training, offered to all employees ensured that he was mentally, emotionally and technically prepared for the job. Today, he is one of the most senior people on the floor.

All in a day's work
As a deputy to the team of 60, one of the first things Srivastava has to do every day is prepare a checklist. Currently, he has seven clients in his kitty. "Every client has a list of agreements that a certain number of items need to be scrutinised in a stipulated time period. I have to ensure, my team meets the expectations," he says. He also has to take a call on matters that his juniors are unsure about. "Sometimes, you aren't sure whether you should approve a profile. It falls in the borderline category. At such time, I'm expected to take a decision," he says. In fact, just hours ago Srivastava made a note of a certain message from a user that sounded "like a sob story". It turns out, that message was sent to over 100 different users. "It was from a man who said he recently lost his family in a car accident, and now he is distressed and in dire straits. It was a call to help. He sounded like a con," he says. It's Srivastava's job then to ensure that the entire floor is notified of this message so that it does not appear again.

Rajeev Srivastava
Rajeev Srivastava

Why manually filter?
Recently, Facebook made headlines for spending nearly a year upping the number of content moderators it employs following accusations that its algorithms alone were incapable of handling the job. "There's only so much that Artificial Intelligence (AI) can do. Of course, automatic systems can root out extreme content before it hits the site, like nudity and abuse, but humans are needed to take a call on more complex matters," says Srivastava.

For instance, if a user uploads a fake profile on a matrimonial site, it's up to a content moderator to scan for discrepancies in information. Because they have access to different kinds of users' personal data like the IP address, location and details of the system, it's easier to detect. As a moderator, he then has to search the Internet,
social media, LinkedIn for any hint on the person. The fact that they have several international clients makes it a tricky job. "Google is the search engine we trust the most. Because I'm an Indian, I will instantly recognise Bollywood celebrities and be able to flag it. But, if it's a dating profile from some other country, and they have misused the image of somebody relatively unknown, it becomes difficult to detect," he says.

When it comes to e-commerce sites, reviews, chats and product descriptions are analysed. "Sometimes, people make escort requests on a product review. So you have to be alert," he says. Having said that, errors occur, he admits. "Not too often, but sometimes you tend to approve a profile which might be fake. We're only human," he says.

Growing numbers
At Srivastava's agency, three million items are scanned in a day. "We have 300 staffers in India and have hired 40 people in Russia," says founder and CEO Suman Howlader, who was one of the first entrepreneurs in India to set up a content moderation company back in 2010. "When I set up the agency, content moderation was an almost unheard of concept. Having a tech background, I realised that the number of bots and spam was on the rise with everything moving online," says Howlader. This realisation has hit home, too, as more and more firms are hiring footsoldiers to rid their websites of objectionable content. Candidates are hired for their sharp analytical skills. In the last 10 years that Howlader has been in the business, the rate of rejections has considerably dropped. "About six years ago it was 40 per cent, and now it's 20 per cent. There's a growing awareness that not everything you write or post will be allowed to slide," he says.

Catch up on all the latest Mumbai news, crime news, current affairs, and also a complete guide on Mumbai from food to things to do and events across the city here. Also download the new mid-day Android and iOS apps to get latest updates

DISCLAIMER: mid-day and its affiliates shall have no liability for any views, thoughts and comments expressed on this article.

Dogs starve to death after NGO staff go on Diwali leave

This website uses cookie or similar technologies, to enhance your browsing experience and provide personalised recommendations. By continuing to use our website, you agree to our Privacy Policy and Cookie Policy. OK