Social media trail of who you are

DeepSense taps into social media posts and online interactions to sketch individual profiles

November 27, 2017 01:24 pm | Updated November 28, 2017 05:10 pm IST

Watch out! All those postings you put out publicly on social media can reveal the type of person you are.

The tweets, LinkedIn posts, photos, videos on Instagram, all hide within them clues of individual personality traits. And an extensive machine analysis of this data randomly left online, can provide a reasonable indication of aspects like whether a person is positive, social, temperamental, considerate to others, and much more.

Social data intelligence company Frrole has developed a state-of-the-art Artificial Intelligence engine, DeepSense, which leverages technologies like Machine Learning and Natural Language Processing, in order to make sense of billions of real-time online social conversations.

The power of machines to learn has grown to such an extent that it’s finding application in segments like employee recruitment and customer support — two areas the three-year-old startup is currently focussed on.

To get an idea of what tweets reveal about an individual, head to https://frrole.ai/deepsense-app, and fill in your or anyone else’s Twitter handle.

Twitter is just one of the social platforms analysed, says Amarpreet Kalkat, Co-founder & CEO. “We also track Instagram, LinkedIn, Klout, Stack Overflow. Only public postings are tracked. For that reason, Facebook is missing because by default it is private by nature.”

A little data reveals a lot

Are we inferring too much about the power of Machine Learning and analytics? “No,” Kalkat says emphatically. “One thing people don’t realise is that how little data is required to start making deductions about you, and probably correct enough.”

By learning about individuals’ interests and personality, employers can understand how they might behave in real life, as an employee — such as, will they be positive, cynical, self-motivated, action-oriented, team players, goal -ocussed. “These are things that recruiters who handle thousands of resumes struggle with. Here, AI will help the recruiter shortlist without having to speak to all of those people,” he says.

For example, an Android developer who regularly posts (publicly) about related topics and engages in discussions on them, will likely be more passionate about the subject than someone who posts and comments about movies. Machine Learning is redefining online customer relations too, which generally is a context-less interaction: the responder might not know who the customer is; whether the person is a he or a she, of 14 years or 40 years, regular customer or a first-timer. This is in stark contrast to the offline customer relations interaction, wherein the responder starts making deductions about the customer the moment the latter walks into the room. “DeepSense is bringing in a layer of intelligence, helping the responder get a better idea of who the customer is,” says Kalkat.

One of the companies which has deployed DeepSense is Freshdesk, a player in the online customer service industry, which is using this technology to help customer service agents. Freshdesk is now getting information about customers in addition to what it already has, and web-derived intelligence. “Now, our users can understand their customers deeper and engage in more customised conversation,” says Ganesh Ram Natarajan, Sr Director, Product Management, Freshdesk.

Auto-learning process

But when AI-based prediction models aren’t 100% accurate, won’t the errors make the process a flawed one? Kalkat says the system has built-in safeguards to take care of that aspect. One is that the AI looks at multiple segments.

“We all leave some footprint somewhere, some time, that we don’t even realise. It’s not just the comments you make, but also the people whom you follow, the likes you mark, the people with whom you are connected, and engage with. This is where Machine Learning comes in.” The algorithm looks at not just explicit signals, but also correlations and patterns.

Another way is the self-correcting loop: the system throws up recommendations based on what it knows about you, and then asks you if what was suggested is right or not. And depending on your answer, the algorithm learns. And that is a continuous process.

0 / 0
Sign in to unlock member-only benefits!
  • Access 10 free stories every month
  • Save stories to read later
  • Access to comment on every story
  • Sign-up/manage your newsletter subscriptions with a single click
  • Get notified by email for early access to discounts & offers on our products
Sign in

Comments

Comments have to be in English, and in full sentences. They cannot be abusive or personal. Please abide by our community guidelines for posting your comments.

We have migrated to a new commenting platform. If you are already a registered user of The Hindu and logged in, you may continue to engage with our articles. If you do not have an account please register and login to post comments. Users can access their older comments by logging into their accounts on Vuukle.