Jinni is nothing like IMDB. You don’t use it to find facts about movies and actors, you use it to find movies to watch — movies that are right for your mood and your taste.

The Jinni team calls their product a semantic discovery engine for movies and TV shows. What does that mean? I searched by entering my mood (rough, humorous), the kind of plot I’m interested in (police investigation) and a period (80s). Jinni recommended movies like 48 Hours, Lethal Weapon and Red heat.

This made me remember how much I enjoyed Nick Nolte in 48 Hours and I select this movie. Now I get several options: I can add a review, add it to my favorites, add it to my wish list or watch it right away. If I want to watch it, I can choose between several ways to rent it or stream it, most of which don’t work outside the US. There are also links to online stores where I can buy the DVD or Blu-Ray disc.

Searching or browsing

My first time on Jinny I was unsure about what to put in the search box - this is a different kind of search. To explore Jinni’s categories, choose the browse option on the front page. These are the choices you get:

• Mood
• Plot
• Genres
• Time/Period
• Place
• Audience
• Praise

Each of these categories have numerous subcategories, all represented by images from well known movies. Jinni is easy to navigate and fun to use.

The Movie Genome Project

Jinni’s recommendations are based on semantic search technology developed through the Movie Genome Project.

The Movie Genome is an ongoing project where the Jinni team works with the Jinni community to map aspects of movies and TV shows. The Movie Genome powers the search, recommendations, taste types and more on Jinni.

The Genome is divided in two: Experience (the mood and tone of the content) and Story (plot elements, structures and more. It also includes external aspects like awards.

The Jinni team maps the “genes” of each title manually. Every movie gets around fifty genes, among thousands of possibilities. This step is manual. The next step uses advanced machine-learning technology: Jinni’s system learns from the manual tagging to begin automated tagging. The collaboration between human and machine creates a level of consistency that creative human taggers can’t reach.

Get an invite

If you love movies you’ll want to try Jinni. It’s fun and you’ll learn a lot. If you are interested in semantic search and cutting edge search technology, you’ll also want to take a look. Jinni is in closed beta, but we have 300 free invites for Pandia readers.

Of related interest

The Theseus image and video search project
Top 5 Semantic Search Engines

2 min intro to Jinni