We are all familiar with recommendation systems. When at Amazon, there is a “Recommended for you” section with items that, based on your previous purchase and browsing experience, Amazon thinks you may be interested in. Netflix too recommends movies that, based upon your previous selections, thinks you may watch. YouTube makes recommendations as well and Facebook dives into your personal profile and selects ads based upon that information.

Whether you are paying attention or not, recommendation systems are on many sites with ads and products. The systems work pretty well. In fact, when you are a bachelor you can become brain-dead basing all your purchases and movie selections solely on what is recommended to you.

But then you get married. Hair creeps into the sink and Sisterhood of the Traveling Pants somehow makes it into your Netflix recommendations. Then, in the year 2022 when you are a dad of a teenage son, you may find an abnormal amount of rap videosĀ  mixed among your moon-scape improvement tutorials.

It is inevitable that most recommender systems will fail to compartmentalize the viewing habits of different members in the household. Setting up a per-user login fixes this, but typically Netflix, Amazon, and iTunes are shared among household members for reasons of practicalities (use of same address or credit card). It works great when recommendations stay black and white, but when they turn grey, it can be a problem (The Joy of Moon Garden Birthing and Nursing, would be a grey recommendation).

So we need a solution to keep users of a shared account separate in order to improve the recommendations of a recommender system.

How about mouse movements? In a Patent applied for by Google: System and method for modulating search relevancy using pointer activity monitoring (7,756,887) it is proposed that mouse movements be used to track what a user is reading on a particular page in order to produce more relevant search results and ads. It was stated that how a user moves his or her mouse over a page is linked to what he or she is reading (how many of you are moving your mouse over the text of this blog post right now?). A great explanation of what this means can be found on the web site SEO by the Sea in Bill Slawski’s blog post titled:
Where you Point Your Mouse May Influence Google Search Rankings, Advertisement Placement, and Oneboxes.

Okay, so mouse movements can tell Google, or any web site for that matter, what a user is reading and can present relevant results. But how does that tie in to multiple users? The answer: Movement rhythm.

If Apple can propose knowing whether a user of an iPhone is the actual owner by monitoring the user’s heart rate and rhythm (everyone’s is slightly different), Netflix, with the right algorithm, should be able to recognize the idiosyncrasies of mouse movement among different members of the household.

On the surface we see a mix of movies from crime drama, Sci-Fi, 80s comedy, chick flix, Tom Hanks and Julia Roberts movies. A standard recommender system may assume crime drama, Sci-Fi, and 80s action flix belong to the man of the house, and chick flix, Tom Hanks, and Julia Roberts belong to the woman of the house. In my house it would be incorrect.

I’m assuming that my wife’s and my mouse-moving habits are slightly different. My mouse movements are probably quick and erratic, hers slower and more methodical. Instead of assuming all searches for Julia Roberts were for the female user, based on mouse movements the site would be able to determine, “this is User A, he may like Conspiracy Theory or Ocean’s Eleven” and not confuse me with User B, who may instead prefer Steel Magnolias or Eat, Pray, Love.

Search for Tom Hanks? Well the user’s mouse movements are quick and erratic, [he] may be interested in The Da Vinci Code, Catch Me if You Can, Green Mile, and Bosom Buddies Season I–um, yes, remember, 80s boy here. Search for Tom Hanks 3 hours later? Well, the user’s mouse movements are now slow and methodical, [she] may be interested in Sleepless in Seattle, You’ve Got Mail and Every Time We Say Goodbye.

Of course mouse movements only work on computers, we’ll need to come up with other ways to capture bio-feedback through remote controls, touch screen phones, etc. but I think we have a good start. Netflix, I’m here to consult for a reasonable fee. Also, I wouldn’t mind free movies and digital content for life.

About Chad Leigh Kluck

I enjoy technology development and management by following new trends, change and disruption, and security. I have a Master of Science in Software Engineering and my hobbies include railroads, history, do-it-yourself projects, writing, and ham radio (K0RRX). More...

I enjoy comments and feedback. Let me know if you have anything to share with me like stories, pictures, or other media. You can also contact me directly via the contact form. I'd love to hear from you!

Leave comments or feedback