All posts tagged: VineSleuth

Big data supports expert wine tasters

In the course of developing software for predicting consumer wine preferences, a Houston-based start up, VineSleuth, shed new light on the abilities of expert wine tasters and the validity of blind tasting assessments. Contrary to popular belief, the company’s VineSleuth metrics, which are based on the work of Chief Science Officer Michael Tompkins and his team, reveal that tasters can consistently identify aroma and flavor characteristics in blind wine evaluations. “We have extensive experimental data which support that expert evaluators have the capacity to precisely identify wine characteristics in blind repeat samples,” said Tompkins whose work spans thirteen years in the field of numerical methods. “During the course of our experiments, our vetted evaluators repeat sample characteristics about 90% of the time,” he says. VineSleuth’s data directly confronts the popular misconception that consistent sensory evaluation of wine is a random occurrence. In developing an algorithm designed to help consumers make wine selections based on personal preference, the company has established a benchmark based on the results of its top-performing tasters (including this author) and intends …

Advertisements

Turning Data into Dollars

A virtual tidal wave of data is fueling the race toward accurately predicting consumer preferences and buying behavior. With intelligent computer systems now gathering information in real time, businesses of all sizes are experiencing a sea change in the way they and their customers use technology. Until very recently, we looked to the past and relied on historical data to identify current trends and to predict consumer behavior. But now we’re looking forward. Information technology has spun us around 180 degrees and given us the ability to use data to extrapolate and make predictions. With a new generation of applications being developed to predict consumer preferences and wine-buying behavior, companies of all sizes are clearly pursuing the promise of big data. “Predication is the nut that everyone is trying to crack,” said data scientist Michael J. Tompkins, cofounder and chief science officer for Houston-based startup VineSleuth. For Tompkins, whose typical project may involve millions of unknowns, wine has its own unique set of challenges. “It’s not as if we have the answer yet,” he said. …