The Click On BBC Radio4 series this week, presented by Simon Cox, talked about predictive technology and how it influences the action on the advertising digital signage industry, the stock market and the health arena.
Cox interviewed Ian Guest, NEC Europe IT solutions, who presented a digital displays with an integrated camera and predictive software that can react to people’s positive or negative response to advertising in real time.
“What we have there is a bit of analytics software that is measuring age, sex, our moods, how long we are in front of the display, this is all integrated into the digital signage display,” explains Guest, “So as you walk towards the screen it will display an ad that is relevant to you. The combination of the screen/camera and analytics software can tell what kind of reaction you had to the advert, how far away are you from it and how long you have been looking at it. You gather the information real time, so you instantly get a feel on whether a campaign is working or not, you can adjust it in real time.”
“But do people like this?” asks Cox.
“People seem to be quite happy to received a message as long as no personal information is requested and captured,” says Guest.
A new technology trend is also available, that predicts not just a person’s mood but the collective sentiment on the market that can affect the rise or fall of shares, by capturing the sentiment on Tweeter and on Blogs.
Studies have found out that Blogs can be used to predict moods and tweets have been used to predict box office sales. Apparently one can also analyse moods in the tweeiverse using a tool called the Google-Profile of Mood States (GPMOS). The GPMOS is an algorithm designed by researchers Johan Bollen and Huina Mao at the School of Informatics and Computing, Indiana Unviersity-Bloomington, United States.
One hedge found company in London has decided to use this tool to make money. The company has launched a £25m fund that makes its investments by analysing its tweeter feeds. The fund uses this information to predict whether the market will go up or down.
According to Bollen’s studies, it might be possible to improve predictions of the market’s gyrations using the sentiment of the thoughts posted on Twitter.
Interviewed by Cox in the radio programme Bollen explains: “We started making this analysis as a way of validating our collective mood measurements. We were expecting that if the market was doing very poorly, people will become rather sad, or angry or anxious and if the market was performing well, the opposite will be the case. What we found was the opposite dynamic, first the mood of people changed and then the markets followed. On the paper we reported an accuracy of 86% on predicting the up and down movements on the Down Jones.”
Market traders can crunch through a lot of data, but at the end of the day it is their intuition, their sentiment on the way the stock might fair that will make them buy or sale. So in a way, this tweeter analytics is reflecting the same dynamics as there are in the stock market. It’s not too surprising that there would be strong correlation.