Controlling connected devices using a relationship graph

1. A system, including:

a processor configured to:

receive a command;

apply a graph model to identify a subset of nodes related to the command, wherein the graph model includes a plurality of nodes that each correspond to a corresponding network connected device and the graph model includes a plurality of edges that specify relationships between the plurality of nodes, and wherein the subset of nodes related to the command is identified including by identifying a specific node of the subset of nodes as being associated with an associated physical area of the command and identifying one or more of the network connected devices associated with the associated physical area including by using the graph model to identify one or more other nodes of the subset of nodes connected to the identified specific node;

activate the subset of nodes in response to the command, wherein activating the subset of nodes includes automatically controlling at least a portion of the network connected devices, using a result of the activation of the subset of nodes in evaluating a plurality of output candidates according to an optimization function, and selecting an output that optimizes the optimization function among the output candidates; and

receive a feedback associated with the output and update one or more reference optimization evaluation values based at least in part on the received feedback; and

a memory coupled with the processor, wherein the memory is configured to provide the processor with instructions.

2. The system of claim 1 , wherein the command is not an explicit user command and the command was detected based on a detection of an environmental property.

3. The system of claim 1 , wherein one or more of the plurality of edges are associated with a dynamic weight value.

4. The system of claim 1 , wherein the graph model is implemented as a sparse table or a sparse matrix.

5. The system of claim 1 , wherein the command is associated with an automation rule.

6. The system of claim 1 , wherein the graph model is hierarchical and includes a plurality of levels.

7. The system of claim 1 , wherein one or more of the plurality of edges identify a physical relationship between one or more of the plurality of nodes.

8. The system of claim 1 , wherein one or more of the plurality of edges identify a functional relationship between one or more of the plurality of nodes.

9. The system of claim 1 , wherein for a particular behavioral activity, one or more of the plurality of edges identify a relationship between one or more of the plurality of nodes.

10. The system of claim 1 , wherein a plurality of different types of edges connect two same nodes of the plurality of nodes.

11. The system of claim 1 , wherein the command is associated with performing a test to verify one or more nodes and edges of the graph model.

12. The system of claim 1 , wherein the command includes a voice command.

13. The system of claim 1 , wherein the command is associated with an automatically discovered automation rule and the automation rule was discovered using the graph model.

14. The system of claim 1 , wherein applying the graph model includes identifying a location of a pet or a human user and a corresponding node of the graph model to the location of the pet or the human user.

15. The system of claim 1 , wherein applying the graph model includes using an automation rule discovered using a different graph model.

16. The system of claim 1 , wherein applying the graph model includes determining nodes associated with a location identified by the command.

17. The system of claim 1 , wherein the subset of nodes related to the command has been identified based on properties associated with one or more edges of the subset of nodes.

18. The system of claim 1 , wherein at least one of the subset of nodes related to the command has been identified based on a relationship of the at least one node with respect to an outdoor environmental property.

19. A method, including:

receiving a command;

using a processor to apply a graph model to identify a subset of nodes related to the command, wherein the graph model includes a plurality of nodes that each correspond to a corresponding network connected device and the graph model includes a plurality of edges that specify relationships between the plurality of nodes, and wherein the subset of nodes related to the command is identified including by identifying a specific node of the subset of nodes as being associated with an associated physical area of the command and identifying one or more of the network connected devices associated with the associated physical area including by using the graph model to identify one or more other nodes of the subset of nodes connected to the identified specific node;

activating the subset of nodes in response to the command, wherein activating the subset of nodes includes automatically controlling at least a portion of the network connected devices, using a result of the activation of the subset of nodes in evaluating a plurality of output candidates according to an optimization function, and selecting an output that optimizes the optimization function among the output candidates; and

receiving a feedback associated with the output and updating one or more reference optimization evaluation values based at least in part on the received feedback.

20. A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:

receiving a command;

applying a graph model to identify a subset of nodes related to the command, wherein the graph model includes a plurality of nodes that each correspond to a corresponding network connected device and the graph model includes a plurality of edges that specify relationships between the plurality of nodes, and wherein the subset of nodes related to the command is identified including by identifying a specific node of the subset of nodes as being associated with an associated physical area of the command and identifying one or more of the network connected devices associated with the associated physical area including by using the graph model to identify one or more other nodes of the subset of nodes connected to the identified specific node;

activating the subset of nodes in response to the command, wherein activating the subset of nodes includes automatically controlling at least a portion of the network connected devices, using a result of the activation of the subset of nodes in evaluating a plurality of output candidates according to an optimization function, and selecting an output that optimizes the optimization function among the output candidates; and

receiving a feedback associated with the output and updating one or more reference optimization evaluation values based at least in part on the received feedback.