Unfortunately,  traditional TV systems (e.g. satellite TV) only allow one-way communication, preventing the use of a centralized recommendation engine. Thus, customers of broadcasting services (DVB-T and DVB-S) still mostly consume a traditional TV experience, where no personalization is provided and searching for appealing content to watch on TV is limited to either browsing the EPG (Electronic Programming Guide) or zapping among hundreds of channels.

The use of recommender systems in broadcasting services would let service providers offer a set of new applications, such as, for example:

• Push VoD. The STB could preload a personalized subset of the on-demand content catalog, where movie selection is tailored to the subscriber’s preferences. The STB internal hard drive has a limited capability and the recommendation technology would allow to optimize the use of this limited resource by storing a selection of movies and programs that are likely to match the subscriber’s tastes and interests.
• EPG Recommendations. The STB could provide recommendations on the EGP on the basis of the user activity history, helping the user to rapidly sift through hundreds of channels and to find interesting TV programs.
• Targeted Advertising. The STB could preload a set of personalized ads and insert them in Linear TV as well as VoD. Delivering the right message to the right customer at the right time is a key factor for advertising effectiveness.

Can we provide personalized recommendations to customers of one-way broadcasting services?

We have recently introduced an approach designed to do just that at the recent Broadcast Engineering Conference (NAB 2012), just held in Las Vegas.

The main issue with broadcast networks is that they do not allow recommendations to be computed server-side, because it is not possible to collect users feedback, ratings and activity. Thus, recommendations must be computed client-side. It’s the user’s STB that is in charge of generating the personalized list of items (e.g. TV programs of VoD movies) to be suggested to the user.

Let’s see how personalized recommendations can be still processed by the STB under such conditions.

Content-based recommendations can be computed in a straightforward way. For example, let’s assume the service provider pushes EPG tags (e.g., genres, featured keywords) together with the TV program to stream. When the user watches a TV show, the STB tracks his activity and updates his preferences, by updating the list of tags related to the program the user just watched. The tags with the largest number of preferences will implicitly represent the user profile and will drive the recommendations of new content. The main challenge with content-based recommendations is the limited computational capability of low-power STBs, but modern STBs provide enough firepower to perform this task with sufficient accuracy.

Content-based systems are however second-best to community-based recommender systems, which are known to provide the most accurate recommendations. Community-based recommendations in broadcast networks represent the real challenge. In fact, such approaches generate recommendations based on patterns and correlations discovered in the users’ activities; roughly speaking, they suggest items that other users with similar tastes “liked” in the past, according to their viewing histories. Broadcast networks do not allow the service provider to collect user preferences in a central location, preventing the recommender system to analyze such preferences and finding out user behavior patterns.

To implement community-based approaches in broadcasting services, we have proposed at NAB to “learn” community patterns on an external, auxiliary domain, where the presence of a feedback channel allows to centrally collect users’ preferences. A mathematical model of the discovered patterns can then be transferred from the auxiliary domain to our target domain - i.e. the broadcast target user population, and eventually broadcast to the STBs through the one-way communication network. STBs can finally generate recommendations locally using the computational model transferred from the auxiliary domain and applying it to the user activity history tracked locally.

The time is ripe for content personalization to customers of broadcasting services too!

Moviri is a veteran exhibitor of the Forum, and we’ve got some great reasons for you to come and visit us this year.

Top 4 Reasons you should visit Moviri at the 2012 IP&TV World Forum

1. We’ve got a newly designed, larger booth (#129), where you’ll be able to rest a bit in one of our comfy, new chairs. Take a break from walking the show and relax in our cool, new space!
2.  Not only is the booth cool, but so are the people who work there! We always enjoy discussing projects, sharing experiences and offering solutions to industry challenges. In fact, ContentWise pioneered the IPTV recommendation engine industry and has been the leader ever since.
3. We’re releasing ContentWise V4.2 with social network data processing. This new version of software will allow you to connect to Facebook, add the social TV experience, refine user profiles and import content metadata available from Facebook and Wikipedia.
4. We’ll be demonstrating A/B testing, a new capability in V4.2 designed to help digital media operators test which approaches are successful with users by comparing a control group with a test group. With this method, operators can measure revenue per user/ARPU, views per user, average time using the service, how much “lift” in revenue recommendations produce, and other performance indicators that would be impossible to measure otherwise.

See you soon in London!

http://www.broadcastprome.com/2012/01/25/helping-viewers-discover-content-and-operators-increase-sales/

It’s a very high-level piece in which I try to paint, in broad strokes, a picture of the current status of recommendations technologies for digital media. The gist of it all is that while there are many different approaches to recommendations and not all work in all contexts, there is no doubt that, once a digital content catalog (tv programs, movies, music, games, etc.) grows beyond the size of a few hundreds to a few thousand items, the need for a discovery solution becomes easy to justify and its implementation makes for real business results.

When the inventory grows beyond a few hundred to a thousand items, the less well-known items become harder to sell. A well-tuned recommendation engine raises the chance of the discovery and consumption of lesser known content. Sales of mid-tail items – typically items ranked 1500 to 3000 in the inventory – rise sharply, while the ‘tail’ of items beyond 3,000 typically contributes 15% of the sales.

Enjoy the read! Many thanks to Vijaya Cherian for publishing the editorial.

People usually have a multitude of interests. When it comes to media and digital content, our interests and preferences can span any number of different heterogeneous genres, moods and styles, not only within any individual media domain, but also certainly across domains. But can we successfully establish a bridge between a user’s preferences in one domain (e.g., movies) and her preferences in another domain (e.g., video games)? To put it simply, does the fact that I like a certain horror movie mean that I am likely to prefer a certain strategy video game?

Media catalog partitioning on ContentWise administrative interface

These questions are certainly attracting the attention of the recommender system scientific community. In the last two years, for instance, the major conference in the field – ACM Conference on Recommender Systems – has hosted a special workshop on “Information Heterogeneity and Fusion“. Similarly, the IEEE International Conference on Data Mining (ICDM) has hosted a workshop on “Mining Multiple Information Sources“.

As summarized in a short survey at the “23rd IEEE International Conference on Tools with Artificial Intelligence“, several works have concluded that transferring knowledge across different domains can effectively increase the performance of recommendation engines based on collaborative filtering. The problem is not trivial at all and user-based approaches founded on assumptions such as “if a group of users have similar tastes about movies they will have similar tastes about video games” are likely to fail. Some advanced solutions – still limited to academic applications – have been proposed. They try to take advantage of hidden relationships among:

• user ratings. Li et al. proposed, for instance, the so called CodeBook in order to share rating patterns across separated domains. (See “Can movies and books collaborate? cross-domain collaborative filtering for sparsity reduction“.
• items. For instance, Cremonesi et al. recently presented a solution for discovering relationships among items belonging to different domains. (See “Cross-domain recommender systems“).

Here’s a simple example.

A recommendation engine that applies a cross-domain collaborative technique might discover that those who like action movies in a certain demographic group are likely to give higher than average ratings to strategy video games. Now, if in the recommendation engine estimation, I have a preference for action movies, I will be recommended more strategy games, and more often, than the average user.  Advanced cross-domain algorithms are also able to discover hidden patterns among multiple domains. For instance, if players who like strategy video games also like role-playing video games, with such techniques I might be recommended both strategy and role-playing video games, just because the recommender system is aware I like action movies.

Although Social TV added value is real, we continue to believe in automated recommender systems very strongly. Here are the reasons:

• Recommender systems can be seen as cold, because they are objective. Their recommendations are the result of thorough analysis of users’ past activity, of their explicit and implicit ratings (e.g. abort a movie stream 10 minutes into it = bad). They are seen as cold, because the key element to improve their acceptance is often missing in the user interface: an explanation. You might not like Di Caprio, however you might be a fan of movies happening in New York therefore Gangs of New York make sense. You may or may not like the explanation, but the system becomes instantly more credible, more understandable and therefore more human.
• Social TV alone won’t solve the long tail issue. Physical and digital commerce requires to be able to move product in the medium and long tail profitably, not only the most popular assets. In fact, the most profitable products may not be the most popular. Read again “Why The Future is selling Less of More”. While Social TV tends to focus on most popular hits, algorithms know how to mix items from the entire catalog and update listings in real-time based on how people react to them.
• Social TV alone won’t solve the serendipity issue. Entertainment works only if it carries some surprise effect at some point. Algorithms know how to handle this as well, while social comments and ratings keep focusing on most popular hits, mostly known to a large section of the population or the network.
• Social TV alone won’t solve the personal taste issue. Being friends-in-life does not mean being friends-in-taste. Algorithms know objectively how to generate suggestions based on user activity tracking and can effectively separate the noise coming from social networks from valuable recommendations.

The solution certainly lies into a combination of both approaches and the technologies that support them, integration into and import from social network streams and recommendation algorithms. Think of it, wouldn’t it be cool to discover who within my social network statistically shares the same taste as I do, so that I can follow his future comments, checkins and recommendations?

More importantly, it would be great to be able, given a service customer base’s preferences, to optimize which content should be in the catalog at any given time in order to maximize average revenue per user and the overall profitability of the content offering itself.

Can we use advanced algorithms to support content planning? Recommender systems may help. In fact, recommendation engines already provide data from which to infer what customers want or don’t want, such as analytics about viewing trends, performance indicators of recommendations effectiveness on items in the catalog and estimation of the users’ interests (e.g. through ratings) on potential content.

In a simple practical application of assisted content planning, for example, a digital music service feeds to the recommendation engine two catalog datasets:

• the dataset representing the live catalog currently available to users and
• the dataset of all licensable content (including tracks and albums not currently available)

The recommendation engine then applies collaborative, semantic and profiling techniques to identify unlicensed tracks that it WOULD recommend to a given user  if they were available to the catalog, thereby giving content planners actionable information to add potentially successful content to the live catalog.

In sum, assisted content planning allows operators to save costs by discarding items users do not find appealing while increasing their satisfaction by acquiring new content the recommender system estimates to match the users’ interests.

Moviri speech was at 10.30 am in room BL 27.1.8 (first floor); at our booth #18 we shown all job opportunities we are offering in our Milan and San Francisco offices.

View all the photos of the day on our Facebook Page

For further details visit the official event website: http://www.careerservice.polimi.it/go/99344

if you have come across this short piece of writing you’ve probably just laid your eyes on one of the announcements with a headline like “BMC buys Neptuny”.

Today BMC Software (NYSE: BMC) acquired Caplan, the capacity management product that recently gained tremendous market traction, analyst appraisal and market success. With this technology acquisition BMC acquired also  Neptuny’s trademark, Neptuny’s co-founder, Fabio Violante, and part of the former Neptuny’s R&D, Sales & Presales teams. This is a tremendous achievement resulting from an exceptional group of people and a business model made of a mix of love for technology, passion, hard work and integrity. I wish the old colleagues and friends to bring (I am sure they will) the same spirit into the BMC team.

This same spirit will continue from now on to live under a new brand and name,Movìri.

Movìri is a relevant player; retains 100% of the clients, 80% of the people and business and will deploy 110% of the passion. We’ll have the same office, the same business partners, the same company structure including the network of subsidiaries in UK and US. We will continue to work on and with Caplan, serving our customers worldwide, as 100% of Caplan’s consulting skills and know-how stays with Moviri. This is a giant leap forward.

More than ever, we are ‘the’ IT optimization experts. We make your applications run faster, use less resources and meet business demands. We enable your infrastructure capacity planning and virtualization initiatives. We help you gain full control over your IT governance processes and assets. We are still the ones striving to blend university research, fresh ideas and delivery ability to serve our customers. The team behind is the same that has served you and other top enterprise customers for more than 10 years.

We look forward to continue to do so.

Yours sincerely

Paolo BOZZOLA