NEMO - NExt generation recommender systems for home Multimedia cOntent

• Unité de recherche: INNOSUISSE
• Numéro de projet: 12134.1;5 PFES-ES
• Titre du projet anglais: NEMO - NExt generation recommender systems for home Multimedia cOntent

Description succincte (Anglais)

NEMO - NExt generation recommender systems for home Multimedia cOntent

Description succincte (Français)

NEMO - NExt generation recommender systems for home Multimedia cOntent

Résumé des résultats (Abstract) (Anglais)

The objective of this project is the development of a next generation recommender systems for home multimedia content, specially designed for the Logitech Revue Box (Google TV). The overall goal is to design an integrated model (denoted NEMO), conceived like an experts system, capable to combine and to exploit different types of information (sources) and different types of (specialized) recommender systems. Project NEMO should deliver a working prototype that incorporates multiple different content type recommendation streams, incorporates the customer's individual usage behavior, and integrates enhancements to the recommendations based on the customer's social network relationships. The most important functionalities of the projected system are: integration of temporal/spatial information as agenda/calendar information (scheduled activities/events, holidays) or user present location (main home, holiday home), integration of social information, extracted by analyzing the social networks of which the user is a member, capability of learning a user profile using the user feedback recorded through the Logitech Revue Box, integration and aggregation of multiple, multi-items recommendation lists from known authority web-based recommender systems.

Résumé des résultats (Abstract) (Français)

The objective of this project is the development of a next generation recommender systems for home multimedia content, specially designed for the Logitech Revue Box (Google TV). The overall goal is to design an integrated model (denoted NEMO), conceived like an experts system, capable to combine and to exploit different types of information (sources) and different types of (specialized) recommender systems. Project NEMO should deliver a working prototype that incorporates multiple different content type recommendation streams, incorporates the customer's individual usage behavior, and integrates enhancements to the recommendations based on the customer's social network relationships. The most important functionalities of the projected system are: integration of temporal/spatial information as agenda/calendar information (scheduled activities/events, holidays) or user present location (main home, holiday home), integration of social information, extracted by analyzing the social networks of which the user is a member, capability of learning a user profile using the user feedback recorded through the Logitech Revue Box, integration and aggregation of multiple, multi-items recommendation lists from known authority web-based recommender systems.