Amid the great occurrences of chronic diseases, considered by some specialists as one of the great health problems of the 21st century, the diabetes stands out, because each patient presents different needs, be they alimentary, physical and/or psychological. Faced with this reality, is necessary that the diabetic patient has access to information that is in accordance with their reality of living with the disease, which will ensure that the information acquired by him is relevant and, in fact, will contribute to his treatment. Thus, it is necessary to know the patient, their reality, their interests and aspects influence your health. Moreover, it is essential that the information who reach the patient are safe and do not provide erroneous data, what can lead to complications in their health. Thus, this work addresses the development of a Social Network, which integrates an informal learning system for health, the MobiLEHealth, use the tools of Semantic Web to collect data of diabetic patients and, in base of their profiles, recommends contents to assist in the treatment. To ensure that this recommended information is secure and comes from trusted sources, was developed together with the Social Network an Expert Intermediate System, which allows a healthcare professional to evaluate the contents and sources that can be recommended to patients. At the end of development, these systems were submitted to a real users test. With this validation, it was realized the relevance that they present, as well as the importance of implementing these environments in the daily life of a diabetic patient and a health professional