Semantic Media and Communications

Kontakt: Stefan Marx, Fikret Sivrikaya

 

The vision of the Semantic Media and Communications cluster is to achieve a service-level merger among heterogeneous network substrates, and to design context-sensitive, personalized and semantically rich multimedia applications over a common framework. Our research and implementation activities closely follow the technologies and open standards in the context of NGN based Telecommunication and IPTV as well as the Semantic Web. Standardizations by committees like 3GPP, Open IPTV Forum, ETSI, IETF enable personalized multimedia content delivery, TV and communication services over next generation networks. We integrate and enhance those services with semantic Web technologies and approaches, especially user profiling, data integration and representation and recommendation techniques, to assist users in their content selection in a world of permanently increasing amount of multimedia content.

The group develops and maintains its own standards-compliant implementations of enabling components in the domain of service control, media delivery, user equipment. These components are utilized and extended continuously for further research.

As a result of our work, a quadruple play application integrating Internet data, Telephony and TV services in fixed and mobile scenarios was developed [1]. It provides communication and entertainment features like VoIP, Instant Messaging, Presence, IPTV, Voting, personalized advertising, and media broadcasting. Semantic approaches are used to provide personalized recommendation and services adapted to the user based on user-preferences, observed user behaviour, content describing metadata, and context information. From user perspective this increases the quality of the IPTV service experience, while a provider benefits from systematic targeting ability (e.g. for advertisement, content, and added services). This demonstrator was deployed into the SerCHo (Service Centric Home) showroom [2] to be available for interested people and to be used for further research.

Core components are utilized to develop additional services and various use cases. For instance, the knowledge of tracking the user served for contribution in the SmartSenior project [3]. The goal is to reduce the social isolation of senior citizens by matching monitored tv history and additional information like hobbies and preferences [4]. A search engine is extended by a recommendation system specialized for this purpose. In contrast, the Smart Media Assistant uses core functionalities of describes recommendation system, but focuses more on using, sharing and recommending media.

Further research concentrates on building a unified infrastructure for semantic services that comprises a data integration layer based on well known ontologies and recommender system on top of it. Subsequent research issues are the integration with publicly available semantic data provided via linked open data (LOD) and communications related information (e.g. presence data, social web applications). A further focus will be the realisation of multiplatform client solutions conform to NGN, IPTV and Web standards utilising those semantic recommendation services which enable a seamless consuming, sharing, searching and accessing of mutlimedia content.

 

[1] Semantic IPTV – http://www.semantic-iptv.de

[2] Service Centric Home (SerCHo) – http://www.sercho.de

[3] SmartSenior – http://www.smart-senior.de

[4] Baris Karatas, Torsten Schmidt, Jérôme Kunegis, Till Plumbaum, Proc. Ambient Assisted Living 2010