Menu
Home
Log in / Register
 
Home arrow Economics arrow Content Delivery Networks. Fundamentals, Design, and Evolution
Source

Data-Driven Recommendation

As one clicks through various items of digital media on an online platform, and as semantic text searches are gathered against your user profile, it becomes possible to infer recommendations for other items that you may appreciate. So as you search sci-fi movies for Star Wars, it is logical to recommend the latest Star Trek video. The business and technology of recommendations is a sector and a topic in its own right. Amazon has built their success on their recommendation engine, and its ability to uplift sales dramatically.

In the same way a VOD provider such as Amazon Prime or Netflix works hard to ensure that as I browse my movies they “learn” my interests and push a queue of things that by and large do interest me and become part of the value I get from my Netflix account.

Of course, VOD services can be badly implemented. I find it tedious that when I search for a product to buy, and buy it, I then spend several weeks seeing offers and deals for whatever it was at a lower, better price. This type of overly eager recommendation system has its backlashes; pushing adult movie recommendations among the kids movie search results, for example ....

Data and the inference of behavior, if used carefully, can work at massive scale. With today's Big Data sources around users, it can be possible to push complimentary second-screen content to users who are engaged in a broadcast channel. A significant attempt to address that market was introduced with Anthony Rose (former CTO of BBC iPlayer) as he launched Zeebox along with the concept of “social television.” Zeebox was the first of many to attempt to merge social media and broadcast. The idea is that most users today are using a tablet or smartphone while they watch TV, and if that experience and engagement can be combined with the TV experience, then two things potentially happen: first, stronger audience engagement, which has value to the show's sponsors, and second, a back-channel of Big Data emerges, helping the producers better understand their audiences.

Success is varied, and seems to be short-lived even when it works, since many of these projects are attached to events or single TV series.

 
Source
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Mathematics
Political science
Philosophy
Psychology
Religion
Sociology
Travel