Abstract: A method including the following steps: receiving user input; resolving a feature of the input using a trainable algorithm, the trainable algorithm being trainable to resolve a feature by application of the algorithm to a dataset including a plurality of labelled dataset entries, the label of each labelled dataset entry describing a feature; wherein the trainable algorithm resolves the features in user input by identifying in the user input a dataset entry labelled with said feature; forming a UI that incorporates the resolved feature; presenting the formed UI; obtaining feedback in relation to the presented UI or a feature thereof; applying the feedback to train the trainable algorithm to resolve features of a UI, wherein feedback for training the trainable algorithm derives from any one or more of, or a combination of: user selection/validation/customisation of features presented to the user and/or user observation.

Abstract: An integrated heating and cooling system for a living space of a building connects together (a) a solar-based heating and cooling system and (b) heating and cooling systems that rely directly or indirectly on fossil fuels that are conventionally used in buildings. The system operates the solar-based heating and cooling system preferentially to heat or cool the living space. Therefore, the system minimizes the use of fossil fuels to heat and cool the living space.

Abstract: An integrated heating and cooling system for a living space of a building connects together (a) a solar-based heating and cooling system and (b) heating and cooling systems that rely directly or indirectly on fossil fuels that are conventionally used in buildings. The system operates the solar-based heating and cooling system preferentially to heat or cool the living space. Therefore, the system minimises the use of fossil fuels to heat and cool the living space.