Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

DataSift embeds machine learning algorithms into its social media analytics platform

Sam Shead | July 20, 2015
The company launched a product in partnership with Facebook in March to help brands see what people are saying about them.

DataSift is aiming to improve its social media analysis platform by integrating machine learning algorithms into it that provide brands with better insights into what people are saying about them over Facebook and Twitter.

The London startup, which counts advertising giant WPP among its customers, launched a new product today called Vedo Intent which provides marketers with "more nuanced insights" while also revealing "customer intent to purchase".

Facebook iPad
DataSift's platform can be used to help brands gain insights into what people are saying about them on social media 

Traditionally companies have relied on an army of beady-eyed analysts to follow and record what's being said about them on social networks, blogs, news articles and discussions but DataSift claims this manual approach to analytics isn't manageable.

"We have a technology platform where we work in partnership with the world's social networks to allow developers, marketing agencies or brands to do analysis of social media content at scale without them having to become masters of big data," said Tim Barker, head of product at DataSift.

"Our DataSift platform can process tens of thousands of items a second and aggregates that to provide insights into audiences and different topics."

Vedo Intent is a 'supercharged' version of DataSift's existing product, Vedo. It's being touted as a feature that allows people across the business, including marketers, customer relations managers and others, to easily build advanced machine learning-based classifiers for their organisation without the need of a PhD.

DataSift claims that Vedo Intent learns in real time as posts are manually classified into categories such as rant, rave, purchase intent or churn. The product then builds a machine learning-based model to first suggest, and then fully automate, the real-time classification of millions of posts to surface insights that previously would have been hidden.

Barker said an airline could use the platform to help them benchmark how their service compares to their rivals' and use that to determine where they're under performing or over performing.

Tens of millions of posts, comments and Tweets are made about companies across social media channels every day. DataSift launched a product with Facebook earlier this year called Topic Data that helps companies identify what people are saying about them on Zuckerberg's platform.

DataSift's platform is available from $1,000 a month for unlimited users.

An increasing number of companies are looking to enhance their existing technologies by making developments in the areas of machine learning and artificial intelligence.

Google, for example, acquired artificial intelligence startup DeepMind for a reported £400 million last year and has integrated it with a number of its existing products, including Image Search and Google+.

 

Sign up for CIO Asia eNewsletters.