November 24, 2013.
Perfumance designed to help shoppers find the right fragrance. Unlike other fragrance apps on the market, Perfumance, through the use of a proprietary hybrid recommendation algorithm, yields meaningful recommendations. Our algorithm combines the methods and advantages of collaborative filtering and content based recommendation. The app’s highly accurate recommendation algorithm was developed along with the expertise of Viktoria Minya, a Paris-based perfumer.
First, the appp asks users basic questions, about style and personality – before moving onto more details, like which fragrance the user wears regularly. For each perfume, Perfumance lists traits and characteristics and sorts them into preferred and unpreferred traits, adding an extra level of precision to the algorithms. Based on these inputs, Perfumance generates a Try List, a list of the top five perfumes the user is most likely to enjoy. The results are given a percentage-based ranking to show how closely they align with the input preferences. As users test and rate new fragrances, they develop a Taste Profile and build a 'Want' list.
The more the app is used, the ‘smarter’ it becomes – and the more precise the recommendations become. Testers have said the app becomes highly accurate after 8-10 ratings. The social features of the app let users share their wish list and give suggestions on what fragrance to buy others as gifts. Users can share their taste profiles with Facebook friends. Perfumance updates its database with the latest fragrance releases, and users can also submit scents. Plus – it includes information about the fragrances top celebrities wear. Read more: