Browse Prior Art Database

Method to dynamically re-order list based on user hesitation and reactions tracked via eye tracking

IP.com Disclosure Number: IPCOM000248583D
Publication Date: 2016-Dec-20
Document File: 3 page(s) / 70K

Publishing Venue

The IP.com Prior Art Database

Abstract

A Method to dynamically re-order lists based on user hesitation and reactions captured via eye tracking

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 48% of the total text.

Method to dynamically re-order list based on user hesitation and reactions tracked via eye tracking

Disclosed is a method to dynamically re-order lists based on user hesitation and reactions captured via eye tracking.

In this age of internet browsing a user typically finds themselves faced with pages and pages of results, firstly when the user searches and then more specifically when they are faced with lists of products or services. Online stores can have hundreds and sometimes thousands of products for users to look through. This can be very time consuming. Even with the use of filters to narrow down the type of product, and the quality of this varies across sites anyway, users can still spend a lot of time sifting through things they are not interested in. Where available users can provide a very narrow search string to help filter items, but this risks eliminating items they care about because sites cannot predict exactly how the items users care about will be described.

When a user is browsing down a long list/pages of products/services, this solution identifies what products/services the user hesitates upon using existing eye tracking technology. Using known criteria about those products, this solution identifies patterns of interest to dynamically reorder and prioritise upcoming pages/content. This is repeated continuously throughout the browsing session to better refine the ordering.

Whilst the user is browsing list a of items, an eye tracking solution is taking the following measurements:

For each item that has been shown from the list, how long did the user1. peer at it?

For each item, how many unique peers were made at it?2. Additionally this solution also maintains:

A list of items that have actually been clicked on.1. These metrics are fed back to the site servers via asynchronous JSON, and

the server stores/updates the measured metrics against the user session within a database.

Item # Total time looked at (s)

Viewed # Clicked # Score

65-65454355 61 1 1 122

11-54869794 203 2 2 1218

54-56463433 1 1 0 1

73-54353664 5 2 0 10

87-43543527 12 1 0 12

A score is generated for each interacted item to provide a numeric indication of the level of user interest in that item. From the above example it is evident that item # '11-54869794' has the greatest level of interested, calculated thusly: (Total time looked at) * (Viewed #) * (Clicked # + 1).

At this point the process extracts the metadata for each listed item from the necessary table.

2

Item # Product Category

Colour Price Size

11-54869794 T-Shirt Black £29.99 Medium

54-56463433 Sunglasses Silver £9.99 -

65-65454355 T-Shirt Black £19.99 Large

87-43543527 T-Shirt Red £39.99 Medium

73-54353664 T-Shirt Gold £249.99 Extra Large

As it is now known to the system that item # '11-54869794' is the most popular item for the user in this session, the system is most interested in the metadata for that item. From this point, a different table is maintained server side against that session. This tab...