As online shopping destinations carry an increasing amount and variety of products, finding what you really want becomes harder. We wanted to find out how the IBM Watson technology can improve the online experience and guide you faster to products that best match your personality and style.
How it works
The Watson personality insights service accepts text input about a person (e.g., free text, Twitter or Facebook feed) and computes a personality profile that ranks that person according to a number of traits. To create the matching service, we computed and stored the profiles for each of the designers we carry. When a customer arrives at our site he/she provides the text needed to compute his/her own personality profile. We compare the profile against the profile of our designers and select the ones that exhibit the most similar characteristics. We then guide the customer towards these products.
Challenges I ran into
To the best of our knowledge, this is one of the first applications of Watson in online shopping. We are analyzing the data we are collecting and expect to make several improvements to the matching model further down the road.
Accomplishments that I'm proud of
A fun project using Expression Engine, PHP and the Watson Bluemix APIs.