Your Movie Wish Is Jinni’s Command
For those of you who are still not satisfied with the current crop of movie recommendation services, you’ll soon have a new choice available to you. Last week, I signed up for the private beta of Jinni and have been pretty impressed so far. Jinni is a new interactive movie rating website that is trying to do for movies, what Pandora has done for music.
While the site doesn’t stream any of the films that they recommend, they do offer convenient links to places where you can find the films online (Netflix, Blockbuster, Hulu, etc.) Apparently, the company has been live for a few months now, but I only just found out about them last week after seeing a review of the service on Read Write Web.
The site includes reviews, photos and even trailers for each film in their database, but their movie filtering software is the real bread and butter. Most of the content you’ll find on their movie description pages is pretty much available on any of the other movie sites, but their “movie genome” information is exclusive to them.
Through a process of human and computer intervention, they’ve categorized every film in their library using information from the movie’s plot, mood, genre, time period, critic reviews, story type, and attitudes. Viewers are then able to filter their search results by using these definitions.
For example, a search for the term bank brings up 134 movies, but if I filter this list by looking only at the “witty” films that include a heist in their plot and are set in the 21st century, I’m able to narrow my search down to just three films, Criminal, Inside Man and High Heel’s and Low Lifes. Since I haven’t seen any of these movies, it’s hard for me to tell how effective this really is, but by narrowing down broad based searches, it does enable me to discover movies that would have gotten lost in the volume of other search results.
On Jinni’s website you can find more information on the actual genome mapping process.
“The starting point of the Movie Genome is manual tagging by our team of film professionals. Each title has around fifty genes, among thousands of possibilities. Then, using advanced machine-learning technology, Jinni’s system learns from the manual tagging to begin automated tagging. This creates a level of consistency that creative human taggers can’t reach – especially important for similarity matches and recommendations, which won’t work unless you compare apples to apples and battles to battles as often as possible. Users who vote on genes, as well as the Jinni team, constantly check and improve the machine tagging.”
After playing around with the site, I was really impressed with the user experience, but I’m still on the fence about whether or not Jinni’s approach is the right way to go. On one hand, by creating “genome” fields around each film’s “DNA”, they’re able to accomplish a lot more with the data, but on the other hand, by restricting rating population to just their staff, it also limits the number of films that they are able to catalog. As an example, if I do a search for the plot Psycho, I get 270 results, but the same search on the user driven site Spout, gives me 509 movies. Now I’d be willing to bet that Jinni’s quality is better then Spout, but by not allowing their users to tag films, they may be giving up quantity through their process. Some people prefer quality over quantity, but I can’t help feeling like they are missing out on the wisdom of the crowds by excluding users from participating in the genome mapping process.
In addition to their movie filtering technology, Jinni also allows you to share more information about your own movie tastes and they provide personalized recommendations. While I haven’t tested the quality of their movie recommendation service yet, I do plan on putting them through my own blind taste test to find out how accurate their ratings really are. In the meantime, if you’re interested in trying the service, feel free to apply to their private beta or you can leave me a comment and I’ll be happy to share one of my invites with the first 10 readers to respond.
Posted on January 26th, 2009 by Davis
Filed under: Disclosure - I own stock in co. mentioned, Media, Netflix, Search, TV, VOD, Web 2.0
Thanks for your thoughtful overview of our service. As you note, we index each title across many parameters (plot, mood, style, etc.) before adding it to our catalog. However, because our system for doing so is automated with human checking (you can read more about it here – http://www.jinni.com/movie-genome.html), we can add new titles quite efficiently and our catalog is growing rapidly.
We always welcome feedback – would be pleased to hear what you think of your recommendations!
Actually, I see that you clearly read the Movie Genome page, so apologies for the redundant link.
Thanks Phoebe, good luck with the site. I look forward to watching it expand.
Dude, you're using Disqus! Dale finally twisted your arm?
His enthusiasm for the service made me figure that I should at least try it. Still too early to tell if I'll keep it, but we'll see.
A more useful, secure, and significant upgrade for *you* would be getting on WordPress 2.7. However, it's a bigger project. And I suspect 2.7.1, with several bug fixes, is due shortly.
Have U seen http://www.movies.monigo.com?