In this paper, we present the design of a mHealth application that is specifically targeted to caregivers and persons with mild to moderate dementia. The result is AREGIVERSPRO-MMD: a tool integrating a broader diagnostic approach, incorporating the live-in family caregiver-patient dyad and considering this dyad as the unit of care. CAREGIVERSPRO-MMD will provide value-added services based on social networks, tailored interventions, clinical strategies and gamification for improving the quality of life for people living with dementia and their caregivers, allowing them to live in the community for as long as possible.
CAREGIVERSPRO-MMD is an EU project funded under the H2020 program.
Sànchez-Marrè, M.; Rodríguez-Roda, I.; Comas, J.; Cortes, U.; Poch, M. Computación y sistemas: revista iberoamericana de computación Vol. 4, num. 1, p. 53-63 Data de publicació: 2000-01 Article en revista
It is clear that nowadays analysis of complex systems is an important handicap for either Statistics, Artificial Intelligence, Information Systems, Data visualization... Describing the structure or obtaining knowledge from complex systems is known as a difficult task. It is innegable that the combination of Data analysis techniques (clustering among them), inductive learning (knowledge-based systems), management of data bases and multidimensional graphical representation must produce benefits on this line.
Facing the automated knowledge discovery of ill-structured domains raises some problems either from a machine learning or clustering point of view. Clustering based on rules (CER) is a methodology developed in  with the aim of finding the structure of ill-structured domains. In our proposal, a combination of clustering and inductive learning is focussed to the problem of finding and interpreting special patterns (or concepts) from large data bases, in order to extract useful knowledge to represent real-world domains, giving better performance than traditional clustering algorithms or knowledge based systems approach.
The scope of this paper is to present the methodology itself as well as to show how CER has several connection points with Knowledge Discovery of Data. Some applications are used to illustrate this ideas.
Using domain knowledge in unsupervised learning has shown to be a useful strategy when the set of examples of a given domain has not an evident structure or presents some level of noise. This background knowledge can be expressed as a set of classification rules and introduced as a semantic bias during the learning process.
In this work we present some experiments on the use of partial domain knowledge with the tool LINNEO+, a conceptual clustering algorithm. The domain knowledge (or domain theory) is used to select a set of examples that will be used to start the learning process, this knowledge has neither to be complete nor consistent. This bias will increase the quality of the final groups and reduce the effect of the order of the examples. Some measures of stability of classification are used.
The improvement of the concepts can be used to enhance and correct
the domain knowledge. A set of heuristics to revise the original domain theory has been experimented, yielding to some interesting results.