This work package investigates and develops a long-term cloud data pattern analytics framework that is capable of providing intelligent feedback towards the coaching application layer. The key objectives are estimating personal mood, social network analytics and context-dependent recommendation for adapting and recommending or automatically switching coaching application strategies on the basis of the huge amount and diversity of historic data streams that are acquired about the contexts, physical, mental and social well-being of individual people but also masses – possibly stored and shared across different social media networks.
Estimating personal mood from social media to adapt coaching application
This task will focus on the development of a social media-based personal mood estimator to implicitly evaluate the exercise and adapt coaching application flow by grounding emotional states acquired through cross correlating data both from sensors on the mobile platform and analysis of social media.
Contextual personalised recommendation to adapt coaching application
This task will focus on the development of a recommender system that generates alternative sport programmes of different lengths for different types of users (e.g. male, female, age groups) that each are heading given their contexts for their own typical goals in terms of physical, mental and social well-being.
Social networks analytics to adapt coaching application
This task will focus on the development of a social networks analytics tool that provides ‘intelligent’ feedback towards the coaching application in the sense that the end-users can be made aware of potentially novel and viable coaching strategies with and/or of others.
In the first phase the basic methods, techniques and modules for adapting the coaching application are developed, whereas in the second phase they are further advanced and corroborated by the large-scale social media network, context and coaching application related data available.