Jinn or djinn are supernatural creatures in Islamic mythology as well as pre-Islamic Arabian mythology. They are mentioned frequently in the Quran and other Islamic texts and inhabit an unseen world called Djinnestan, another universe beyond the known universe. The Quran says that the jinn are made of a smokeless and "scorching fire", but are also physical in nature, being able to interact in a tactile manner with people and objects and likewise be acted upon. The jinn, humans and angels make up the three known sapient creations of God. Like human beings, the jinn can be good, evil, or neutrally benevolent and hence have free will like humans and unlike angels. The shaytan jinn are the analogue of demons in Christian tradition, but the jinn are not angels and the Quran draws a clear distinction between the two creations. The Quran states in surat Al-Kahf , Ayah 50, that Iblis is one of the jinn. Wikipedia.
News Article | April 13, 2015
Jinni said AT&T is using its recommendation engine to offer U-verse TV video-on-demand subscribers a personalized program guide. The deal, announced Wednesday, is one of Jinni’s biggest since it was founded in 2008. The Tel Aviv-based company also licenses its semantic discovery product to Comcast, VUDU and Microsoft for use on its Xbox gaming consoles. International customers include Belgium’s Belgacom, Spain’s Prisa TV and SingTel in Singapore. Jinni’s recommendation engine is unique in the way it indexes movies and TV shows with “descriptive gene tags” such as “clever heist,” “surprising twist” and “offbeat.” When combined with anonymous data that details the viewing habits of U-verse subscribers, it could allow AT&T to recommend that subscribers who it believes are in the mood for "rough criminal heroes" watch CBS's Hawaii Five-0, 22 Bullets or Gangster Squad. Jinni said AT&T is the largest customer of Ericsson’s Mediaroom platform to deploy its recommendation engine in the United States. In October 2012 – nearly a year before Microsoft sold Mediaroom to Ericsson – Jinni rival ThinkAnalytics said that Microsoft had added its recommendation engine to its Mediaroom IPTV partner program. AT&T is using Jinni on every set-top deployed in U-verse homes, Jinni said. “Working with AT&T U-Verse is a major milestone toward realizing our vision of making the video discovery experience an intuitive, enjoyable part of watching TV,” CEO Yosi Glick said in Tuesday’s announcement. More
News Article | December 1, 2014
Walmart’s Vudu service has finally introduced Jinni’s search and discovery system into its user interface for all Vudu users. Jinni announced that Vudu was a customer at the beginning of this year, along with Time Warner Cable, Bouygues Telecom, C More Entertainment (formerly Canal+), Prisa, and SingTel. Jinni takes what it calls a semantic approach to categorizing content. The result, the company claims, is the ability to provide mood-based search and taste-based recommendations, leading to a higher level of personalization. Jinni, like many other search and discovery specialists, goes beyond those obvious categories; it says it characterizes content by mood, style, plot, setting and more. Netflix, for example, has created highly sophisticated algorithms for its search and discovery process, categorizing movies not only by actors, directors, and genres, but also by any number of other characteristics. But Netflix viewers are still generally limited to searching by top level categories such as title, actor, genre. Jinni also differentiates itself from Netflix by pointing out that its categorization process is automated while Netflix uses a manual tagging process. The challenge appears to be finding clients who agree that giving viewers more search options is a useful thing to do. Vudu is adding features that other services, such as Hulu, introduced some time ago, such as creating a page that displays all the shows any particular viewer is still in the middle of watching, and automatically queueing up the next installment of episodic content. Vudu, with Jinni, is going further than many other video services by letting viewers select sub-genres (action comedies, buddy comedies, etc.), or all movies by plot (heros, romance, time travel, etc.), or mood, or years (the ‘30s, the ‘70s, etc.) Yosi Glick, Jinni Co-Founder & CEO added, “We are proud to have a customer like VUDU adopt our solution using our full discovery feature set. This is further validation of our semantic approach we have built over the years which has now become the dominant trend by the leading players.”
News Article | April 13, 2015
There aren't many solo players left delivering TV content recommendation engines, so maybe it's no surprise that one of the big ones is now looking to expand into new markets. Jinni announced this week that it's getting into the ad tech business. The move by Jinni Media Ltd. doesn't deviate very far from the company's technology roots. Jinni said it will use the same Entertainment Genome powering its recommendation engine to help TV and movie studios target advertising on the web and mobile platforms. The proprietary semantic technology matches content inventory with individual user tastes. Repurposed for the ad business, the technology will provide a Demand Side Platform (DSP) for entertainment advertisers that want to carry their promotions over to an online environment. In addition to TV and movie studios, Jinni's new solution is designed for programmers looking to promote content in video-on-demand libraries. Jinni said the product has already been tested in commercial pilots with major Hollywood studios, and that the results showed significant improvement over traditional demographic-based ad targeting. When it comes to entertainment," declared Jinni Co-Founder and CEO Yosi Glick, "the truth is that totally different demographics often share the same tastes in movies and TV shows. Therefore, online advertising of movies and TV shows needs a fresh approach. With our proprietary moviegoer and TV viewer database and our mature semantic technology to understand user tastes, we can now provide entertainment brands with a much more efficient way to buy their media and target their relevant audiences. In the service provider world, Jinni already has an impressive line-up of customers. Big names from the list include AT&T Inc. (NYSE: T), Comcast Corp. (Nasdaq: CMCSA, CMCSK), Microsoft Corp. (Nasdaq: MSFT) and Time Warner Cable Inc. (NYSE: TWC). (See AT&T Jumps On Board With Jinni.)
News Article | September 5, 2014
Even though it’s had $1.5 million in the coffers via a seed investment several years ago, Movli has remained an ‘under the radar’ startup for most people since its inception in 2011. The Tel Aviv-based startup has recently launched its Web-based portal for film-lovers, as it looks to “transform the world’s movie-watching experience,” as the company puts it. The space it’s entering is pretty busy as things stand – there’s the long-established Amazon-owned IMDb for starters, while Jinni serves as a ‘semantic discovery engine’ designed to help you find ideas for movies to watch. Then there’s the lesser-known Letterboxd, a social network for movie buffs. Movli is looking to centralize the three main activities it believes film fans are looking for online: social interaction around movies, accessing movie information, and getting movie recommendations. There is, of course, one glaring ommission here – film fans generally like to watch films too. But we’ll get to that later. For the most part Movli wants to recommend you films, while simultaneously letting you interact with like-minded people and delivering access to interesting factoids. Movli uses the information you put in, including movie-ratings and watchlists, to build what it calls a “personalized genome” of your tastes. It then promises to serve up an accurate prediction of how much you’ll enjoy a certain film, which is given via its ‘Affinity Meter’. Here’s a quick peek at how it works. While you can browse without signing-in, its recommendation engine is based on your input – so it helps if you elect to not remain anonymous. For this, you will have to use either your Facebook, Twitter or Google+ account – there’s no email option I’m afraid. Next, you will see a row of tabs along the top – ‘All’, ‘New’, ‘Upcoming’, and ‘Free’, which are all self-explanatory (we’ll get to ‘Free’ later). When you hover your mouse over a movie, you indicate if you’ve seen a movie. If ‘Yes’, you give it a score out of 10, and this is added to your ‘Watched’ list. If you haven’t seen it, you can give it a priority rating of high-to-low, and it’s added to your ‘Watchlist’. Alternatively, if Star Trek really doesn’t float your boat, you can banish it to your ‘Blacklist’ forever. Each movie listing sports a brief description, with further taps revealing the cast, key quotes, locations, photos and soundtrack details. But it’s the filters that may prove particularly useful. Not only can you stipulate a decade, genre and rating (from IMDb or Movli itself), but you can focus on a film’s runtime (if you have a short attention span), and even if it’s won awards. For example, you could search for Oscar-winning comedies made in the 1980s, that are less than two hours long. Over time, Movli learns what you like, and you can revisit your various lists in your profile. Moreover, when you follow someone or view their profile, their own tastes are fed in to you recommendations, but tweaked to your own tastes. For example, movies you’ve already watched, or ones that don’t match your specified tastes, are omitted. Though there are direct links to buy/rent movies elsewhere (e.g. iTunes and Amazon), the intriguing little tab called ‘Free’ actually pulls together movies that are freely available to watch online, mostly via YouTube. You’re not going to find big blockbusters on here, but there are few decent classics, oldies and low-budget horror flicks available. Movli taps crowdsourced metadata platform Freebase and Wikipedia for most of its database, with a handful of additional APIs used (including YouTube) to “fill in the gaps.” There’s little question that Movli has been nicely designed, but its focus on ‘Web’ over mobile at a time when increasingly more people access the internet almost-excslusively from their pocket rocket, is a chink in its chain. Yes, Movli works well enough on your smartphone via the Web, but it does lack the usabilty you’ve come to expect from a fully native app. Pinch-to-zoom is fine ‘n all, but it’s not the same as using an app that’s been built specifically for the device that’s in your hand, or if it’s been properly mobile optimized. Still, as a first public iteration, it’s a good start and Movli does provide useful features for filtering through the cacophoous crackle that is the world’s movie library, to unearth something you might just like.
News Article | February 27, 2015
I live in Boston, so you'll excuse me if I've been a bit surly lately. Every parking spot is marked with some jamoke's busted-up lawn chair, giant icicles are tumbling from rooftops like thunderbolts thrown by Zeus, and the mass transit system is pokier than a kiddie train at an amusement park. Except far less reliable. Anyway, here are some apps to help get you through the rest of the winter. Leave me alone. Try Jetsetter (iOS, Web). This last-minute travel-deals site is chock-full of discounted hotel rooms at fancy places around the world. The deals run from so-so to stellar, and include some exclusives that can top out at 60% off regular prices. The iOS app leverages your camera to scan your credit card so you can book a trip in less than a minute. Dangerous. Make sure to load up the Skiresort.info app (Android, iOS). Believe it or believe it, getting 700 feet of snow dumped on top of the Northeast has made for some pretty great skiing out here. Skiresort.info has the lowdown on about 5,000 resorts around the world, and can pinpoint the closest one to you. You can see how many runs are open (hint: all of them, if you're in the Northeast), view zoomable trail maps, look up ticket prices, and even view live video feeds from various resorts. Quantify your feelings with Moodlytics (Android, iOS). My mood has been totally fine, so I don't need something that can track more than 40 variations, complete with charts and calendars and suggestions for how to improve said mood. I don't need to set goals in order to reach an optimum number for my own personal Happiness Index. I don't need to export the data to a PDF file so I can share it with my therapist. I'm fine! Work out at home with Virtual Trainer Bodyweight (Android, iOS). As its name suggests, you use your own body weight to perform exercises from the comfort of your own home. The app features instructional videos for 46 exercises that you can cobble together into a custom workout routine. You'll earn points for each exercise you complete, and you can target specific muscle groups. Take Jinni (iOS, Web) for a spin. You tell it what you like, and it'll recommend stuff to watch on TV and on Amazon, Netflix, Hulu Plus, and iTunes. Jinni's content recommendation engine runs deeper than most, tracking content based on things like mood, plot, place, style, type of humor, story tempo, levity, and a whole bunch of other markers. Take some time to train it, and you'll be pleasantly surprised by its recommendations.
News Article | January 6, 2014
Tel Aviv, Israel-based semantic video discovery technology outfit Jinni is today announcing a fairly big deal that was struck with American video streaming, analytics and monetization company Ooyala. Ooyala, flush with $43 million in fresh funding, will integrate its machine-learning big data analytics systems with Jinni’s semantic discovery to deliver what it deems a “new level of video personalization” for all screens. The service will be available to media companies, broadcasters and pay TV operators across multiple regions, but not just yet: Ooyala and Jinni expect to have joint services in pre-release with select customers in the first half of 2014, with broader availability in the second half of this year. In PR lingo, the joint service sounds something like this: “Jinni and Ooyala will work together to develop and deploy a completely new level of machine learning powered by semantic discovery that will allow TV providers to tailor programming and video viewing experiences to each individual user at a very granular and powerful level; offerings such as personalized channels, custom programming guides, mood-based browsing and search, and viewer recommendations for both live and VoD content across vast catalogs of content. All designed to provide users with easy, intuitive access to content of all types.” And here are some ‘conceptual mockups’ (i.e. what it could end up looking like): The reasons this is kind of a big deal for Jinni, is that Mountain View, California-based Ooyala now collects more than 2 billion events daily from hundreds of millions of consumers watching videos. Ooyala, which counts the likes of ESPN, Bloomberg, NBC Universal, Telegraph Media Group and The Washingtong Post among its customers, can analyze all that data to detect trending content and help operators deliver more personalized viewing experience across screen types. As Jinni co-founder and CEO Yosi Glick puts it: Having followed Jinni for quite a few years now, I’m not surprised the company manages to score deals like this; I’m actually amazed they haven’t yet been acquired, if only for its technology. The Israeli company created what it dubs ‘Entertainment Genome’, which powers its cross-catalog search and taste and mood-based video recommendation technology. Essentially, Entertainment Genome is Jinni’s ongoing attempt to map more aspects of movies, TV shows and semi-professional videos than anyone has done before. It attempts to do so by going beyond your usual categorizations such as keywords and genres, and adding thousands more ‘genes’ that are assigned to each title to describe mood, style, plot, setting and much more. Jinni provides content discovery and personalized recommendations to pay TV and OTT content providers such as Xfinity, Xbox, VUDU, Bouygues and SingTel. Separately, the company is today announcing new business deals with Spanish CANAL+ /YOMVI network from local media conglomerate Prisa, and Canada’s TELUS. The partnership with Prisa is particularly notable, as it marks the first Spanish-language set-top box deployment of the (multi-lingual) Jinni solution. CANAL+ is the leading digital entertainment platform in Spain, encompassing a broad family of channels. Jinni’s platform has also been implemented on TELUS set-top-boxes for TV and movie content recommendations on Optik TV, the Canadian telecom company’s IPTV-based television service.
Jinni | Date: 2013-09-03
Disclosed are systems, apparatuses, circuits, methods and computer executable code sets for generating and providing hybrid content recommendations. One or more recommendation engines are collaboratively arranged based on the conditions of a recommendation request. The collaborative recommendation engine arrangement is used for generating a set of content recommendations for the request.
Jinni | Date: 2013-04-28
Disclosed are systems, apparatuses, circuits, methods and computer executable code sets for generating and providing content recommendations to match the tastes and preferences of a group of users. a Recommendation Engine is used for generating two or more individual content recommendation sets for each of the members in the user group. A Recommendation Aggregation Module is used for adding and combining the individual content recommendation sets into an aggregated recommendation set. a Recommendation Selection Module is used for selecting at least a subset of the content items in the aggregated recommendation set for inclusion in a content recommendation result set. A Profile Engine is used for building individual group users profiles from which a merged group profile is constructed, or for building a single joint group profile based on inputs from multiple group users.
Jinni | Date: 2013-04-28
Disclosed are systems, apparatuses, circuits, methods and computer executable code sets for assessing the relevance and impact of a secondary content, such as an advertisement, presented within primary content. A first assessment module is used for assessing a taste profile based relevance and/or impact of the primary content, and/or the secondary content, on the target viewer. A second assessment module is used for assessing a correlation level between the primary content and/or segments thereof, and the secondary content and/or segments thereof.
Jinni | Date: 2013-02-20
Disclosed are systems, apparatuses, circuits and methods for extrapolating meaning from vocalized speech or otherwise obtained text. Speech of a speaking user is sampled and digitized, the digitized speech is converted into a text stream, the text stream derived from speech or otherwise obtained is analyzed syntactically and semantically, a knowledgebase in the specific context domain of the text stream is utilized to construct one or more semantic/syntactic domain specific query analysis constrains/rule-sets, and a Domain Specific Knowledgebase Query (DSKQ) or set of queries is built at least partially based on the domain specific query analysis constrains/rule-sets.