Abstract:

The fast development of software for medicine caused the arise of new methods of learning about illnesses and complications which appear during pregnancies. The curriculum has to adapt to the student's skills, knowledge for them to reach their potential. Learning analytics provides answers and possible solutions for the pregnant women who come to be monitored and treated. The measurement, gathering, analysis and reporting of biological parameters help the doctors and their students to understand and to optimize the healthcare learning, as well as to improve the applied treatments. The proposed tool combines descriptions, diagnostics, predictions and prescriptions in order to obtain optimal results.

The online learning experiences bring evolution and expansion of knowledge for the medical students who will become the future generations of doctors. The system uses an ontology to build the learner profile regarding the notions about illnesses and complications of pregnant women. The recommender algorithm takes into consideration the medical students who did well while learning from the course where the mandatory notions have to be known. The mandatory notions that appear inside the ontology are defined in an RDF file. The notions are linked using RDF triples that are created between a subject, an object and a predicate. The information about the other consulted materials which are added by the doctors are also taken into consideration when doing the reading recommendations based on the new created triples. SPARQL queries are used for querying the data depending on the ontology reasoning and on the defined rules that satisfy a certain condition. The learning analytics software takes into consideration implicit and explicit metrics. The implicit metric gives the patterns of the user's behavior, without being aware of it. The explicit metric is triggered by the notions which are checked based on validated knowledge coming from the medical staff. The quality of learning is improved by analyzing the described metrics and using the automatic recommender in a time where it is difficult to choose out of the many resources and courses that are available online. The dashboard of the students provides them information to understand the immediate actions which should be undertaken for studying pregnancy complications.

Keywords:

Ontology, machine learning, recommender system, pattern matching.