Jinni Taste Engine for TV and movies envisioned as a Google TV app

Wondering what an app built for Google TV looks like? Jinni showed out at the Cable Labs winter conference with its metadata based search engine, and now it's letting people see what it could do on a Google TV device. It works either through "semantic search" based on understanding queries like "smart love story" and also pulls up social recommendations and personalized recommendations once it's stalked gotten to know you. as is all the rage these days, it can pull in content from several online sources like Netflix, Hulu and iTunes. It compares well to the TotalGuide from Rovi mentioned in our roundup -- we'll see what hits the big screens first, and through what delivery method.

Show full PR text

Jinni is the first Taste Engine for movies and TV shows

Jinni's award-winning guide helps users choose what to watch next amid the abundance of available content (live TV channels, VOD, DVR, and internet services like Netflix, Amazon, and Hulu). Based on innovative semantic technology, Jinni filters the universe of content through the lens of the user's personal moods and tastes.

As a vertical engine specializing in premium video, Jinni achieves a deep understanding of content that horizontal search engines cannot. Jinni powers intuitive discovery tools that allow "couch potatoes" everywhere to easily find what they feel like watching from the comfort of the living room.

Jinni operates a popular destination website (nearing 1 million visitors each month) and offers API-based solutions for Internet content providers and TV operators.

The Technology
The Jinni service is powered by our proprietary Movie Genome, containing several thousand genes. These are assigned to each title to describe plot, mood, style, setting, soundtrack and more – a rich alternative to the usual genre language. New titles are indexed by analyzing user reviews and metadata, using innovative Natural Language Processing technology. This automated process offers efficiency, consistency, and a diversity of viewpoints from incorporating many user reviews.

· Semantic Search: Type "smart love story" or "funny action with a surprise twist" in the Jinni search box – and get real results.
· Personalized Recommendations: Jinni generates recommendations based on a unique model of each person's taste.
· Entertainment Personality: A one-of-a-kind model of each user's tastes, or Entertainment Personality, powers all personalized features from Jinni. The model acknowledges that people usually enjoy a variety of content depending on mood and context.
· Social Recommendations: Jinni helps users get relevant social recommendations by identifying people with shared tastes ("neighbors") and displaying friends' recommendations.
· Integration with leading content providers: enables seamless discovery over the catalogs of Netflix, Hulu, iTunes and more. Jinni for Google TV would also enable discovery over live TV and VOD catalogs.