IBM Watson Trend - let the computer tell you what’s hot this holiday
Tags :IBM Watson ibmwatson analytics
IBM Watson Trend is an application and website that reads the tone of the web to tell you what items are hot for shoppers this 2015 Christmas.
At IBM, we started dreaming of harnessing the collective knowledge of billions of people on the planet to help shoppers easily make more informed decisions.
There is both an Android and iOS application for IBM Watson Trend, but the website is quick and easy to use. Even the mobile web version is easy enough. I found it interesting because most of the items were things advertisers and retailers were telling us what is hot and on sale. That means people tend to buy those sale items and then talk about them. It is tough to really dig into what any type of person would want and hope that it is a trend that can be found.
Prediction also seems to be a big part of IBM Watson Trend. It builds a graph to help predict if something will be around for the next few weeks as well as the past few months
Predictive technologies help Watson understand how long a trend might last. You can tell if a particular phenomenon is a fleeting fad or built to last from the trend graph, which shows the trend's strength over the past three months, but also predicts how strong it will be for the next three weeks.
I picked a random kids toy and saw they were pulling review comments from certain sites on the toys of that brand. So this was not just a trend for a certain toy/device but for an entire brand and line of spinoffs like video games, bikes and more. The shopping link for takes you to the Google shopping tab and pre fills in the name to show you results and where you might find it. I am unsure if there are any affiliate links for IBM behind the scenes.
The Samsung TV line is a great example. The IBM Watson Trend is for the entire brand line, not a certain model or feature. Now for some very specific items there was a definite trend to see. And it was interesting to see the projected high and low mention trend analysis which looked more like some average to me, but I am not Watson.
Does this look useful for you this shopping season?
blog comments powered by Disqus
On Wednesday, December 9th, 2015 by Chris Miller
