![]() | ![]() |
We will address this problem by creating an open source software framework to dynamically personalize the delivery and consumption of multimedia. MyMedia will tame growing volume of content streams by combining them and allow users to sip from a single manageable stream of the most personally relevant content. The project will pioneer the integration of multiple, content catalogues and recommender algorithms in a single system. The project delivers an open source software framework which allows researchers and potential commercial exploiters outside of the consortium to easily plug-in and experiment with new recommender algorithms and content sources. This simplifies take-up outside the consortium and creates an even wider impact. The project will evaluate resulting theoretical user models through a set of lab analysis and field trials. The framework will also be evaluated on multiple trial platforms and will be language agnostic. To understand cultural differences field trials will be conducted in multiple countries and languages. MyMedia will innovate by enabling the creation of media-centric social networks that leverage user generated metadata such as tags and explore other possibilities focused on improving the end user experience such as automatically generated content metadata enrichment. It will pave the way to sharing preferences and recommendation results, in a privacy respecting manner, within communities. The contribution of the European Microsoft Innovation Center (EMIC): The European Microsoft Innovation Center (EMIC) leads this project and is responsible for the architecture and implementation of the Recommender Framework: an SDK that includes a programming framework, sample code and tools for developing domain unspecific recommender systems. Based on the experience gained from previous recommender systems, EMIC extracted and abstracted the common components and provided additional solutions to complex problems that nearly all recommender systems need to solve. It enables the developer of a recommender solution to concentrate on the business logic and not care about complex software development problems like catalog updates during runtime, item versions, event flow, dynamic object referencing etc. In addition, the European Microsoft Innovation Center (EMIC) is also work package leader for one of the field trials.
For more information, visit the project’s Web site at: |