Nowadays many users can be involved willingly and monitored in a more or less intrusive manner with data collected in their homes. This data can be collected from mobile and wearable devices (such as smart phone and smart watch), from sensors disseminated in the home (such as motion detectors and contact switches) and from self-reported information systems (using paper based or web-based ecological momentary assessment techniques). This data can then be useful to characterize the activity, the health and the well-being of the involved person. Enabling people suffering (such as elderly people, people with physical disabilities and people with mental-health condition) to stay in their home as long as possible in good condition is an important challenge for many countries. Firstly, most of people would rather to continue to live in their own home rather than move to a nursing-home or an hospital, secondly the solutions enabling staying at home are also usually cheaper for the society. As a consequence Smart Home and Ambient Assisted Living (SHAAL) systems gain more and more attention. SHAAL systems use information and communication technologies in a person's daily living environment to enable them to stay active longer, remain socially connected and live independently. SHAAL's research covers a wide range of topics. This talk will review the main aspects of SHAAL systems from the data mining point of view. Several case studies will be illustrated. In particular it will emphasize the key role of activity learning and behavior understanding.