Archives for the month of: September, 2011

After my last post about neural networks and movie sales predictors, I looked up the original New Yorker article. Gladwell intros with a quote by philosopher David Hume:

“Beauty is no quality in things themselves: it exists merely in the mind which contemplates them.”

Then the article contrasts this by laying out the evidence that there DO seem to be some standardized evaluators of what human beings find appealing in music and in movies (and websites?). What about books? The canon of literature?

The High Priestess

The High Priestess card from the popular Rider-Waite Tarot deck represents what we don't know or can't define; in other words, "whitespace".

Visually-speaking, I think movies are a fast-moving collection of symbols that our subconscious minds react to. Yep, that’s right, I’m referring to Jung’s definition of archetypes. The symbol of “mother” or “teacher” are universal symbols that we as human beings seem to process the same way, independent of culture or background or upbringing. Not to get too esoteric, but these same ideas of archetypes are the basis for the Tarot deck.

One method of choosing (or predicting) user experience is to create personas for the user. Let’s say I’m building a new site for my personal resume (which I am). Suppose Steve, a middle-aged manager of a marketing department at a medium-sized company is looking for a web developer (or UX/UI designer). Steve knows relatively little about how developers work, but he does know what he finds aesthetically pleasing. My job in building this new site is to please Steve’s eye, but also help him find the information he needs to determine whether I’d be a good candidate for the job. That would include my contact info, easily clicked samples of my code and designs, and my work history (so he knows I’m experienced).

So what if there were methods of mapping these archetypes and emotionally-charged information for website users? How would you use symbols to increase users and/or website sales? It’s a bit Mad Men, isn’t it?


I recently heard about a company that uses neural networks and software to predict movie ticket sales based on words in the movie title and plot synopsis.

This has got to be where most things are headed. My initial thought is, this is exactly what’s missing from website development! Granted, people use SEO methods and meta keywords to improve their Google search rankings, but what if there was a method of analyzing the design elements and user experience of a website as a predictor of website a) sales b) returning users and c) customer-initiated viral marketing?

That would be a great way to sell design AND improve user experience, and it could be industry-specific. Perhaps one design element works better with online clothing retailers than it does with luxury car sites, for instance.

Glad I signed up for that artificial intelligence class at Stanford!

Earlier today someone asked me if I could find out in which cities the Formula 1 races are being held for Japan, South Korea and Hungary…since I am normally incredibly fast at researching tidbits like this, I thought, no problem, just a second. How wrong I was.

I’d like to preface this by saying that I am probably not the intended user for this site, but that doesn’t mean the information shouldn’t readily be available. That’s what usability and user experience (and GOOD design) are about: creating the BEST experience for ANYONE who encounters a site, no matter what their gender, age or sports preferences are.

Immediately I searched for the official Formula 1 site; where to find this sort of information? After a brief scan of the top menu bar, I clicked on “Tickets & Travel”, which brought up a table of countries:

Oh, there’s Japan! I clicked on the flag, and Japan was highlighted. But what city are the Japan Formula 1 races in? My immediate instinct was to again click on the text “Japan”, in the hopes that this would link me to the information I wanted.¬† But nothing happened!

Eventually after scrolling, my eyes scanned the bottom of the page where it read, “Book Now“, which took me to a separate booking page. All right, I was sure this would give me what I need. Who would book a trip to Japan without knowing what city they are visiting?

The next page header reads: “2- or 3-Day Pass to the 2011 FORMULA 1 JAPANESE GRAND PRIX (Suzuka)” with a picture and then “Tokyo” next to the header:

But I am immediately suspicious. What is “Suzuka”? Is that a subdivision within Tokyo? Or another city? I scroll down some more. And scroll and scroll and scroll.

FINALLY, at the bottom of the page I see this:

The Suzuka circuit is located about 37 miles (60 kilometres) south of Nagoya. The most convenient airports to use are:

  • Nagoya: 42 miles (68 kilometres) north
  • Osaka Kansai: 102 miles (164 kilometres) south west
  • Tokyo: 250 miles (400 kilometres) north east

Wow, Tokyo is like 250 MILES from the race track! What if I hadn’t known that, and booked a hotel in Tokyo too? Even Nagoya is some distance away. I experience the same frustration with finding the nearest cities for Formula 1 in South Korea and Hungary as well.

Eventually I find a page listing the calendar for the 2012 races. Though the city or location is in parentheses next to the race, it’s still not immediately clear where the race location is.

How about a geographic map of ALL the races together, instead of just individual race maps, huh Formula 1? The tix are pricey enough anyway…

%d bloggers like this: