Clustering devices in an internet of things (‘IoT’)
View Patent ↗Clustering devices in an Internet of Things (‘IoT’), including: receiving, by a device clustering module, a characteristic set for a device, wherein the characteristic set specifies one or more device attributes and an attribute value for each device attribute; clustering, by the device clustering module, the device into an attribute level cluster based on the one or more device attributes specified in the characteristic set for the device; and clustering, by the device clustering module, the device into a value level cluster based on the attribute value for each device attribute, wherein the value level cluster is a subset of the attribute level cluster.
1. A method of clustering devices in an Internet of Things (‘IoT’), the method comprising:
receiving, by a device clustering module, a characteristic set for a device, wherein the characteristic set specifies one or more device attributes and an attribute value for each device attribute;
clustering, by the device clustering module, the device into an attribute level cluster based on the one or more device attributes specified in the characteristic set for the device; and
clustering, by the device clustering module, the device into a value level cluster based on the attribute value for each device attribute, wherein the value level cluster is a subset of the attribute level cluster.
2. The method of claim 1 wherein the characteristic set is specified as an extensible markup language (‘XML’) structured document.
3. The method of claim 1 wherein clustering the device into an attribute level cluster based on the one or more device attributes specified in the characteristic set for the device further comprises:
calculating, by the device clustering module, a commonality index between the device and a cluster representative for each attribute level cluster, wherein the cluster representative specifies a union of all attributes of each device in the attribute level cluster; and
assigning, by the device clustering module, the device to the attribute level cluster with the highest commonality index.
4. The method of claim 3 wherein calculating a commonality index between the device and a cluster representative for each attribute level cluster further comprises:
determining, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative;
determining, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are not attributes of the cluster representative; and
dividing, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative by the sum of the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative and the number of device attributes specified in the characteristic set for the device that are not attributes of the cluster representative.
5. The method of claim 1 wherein clustering the device into a value level cluster based on the attribute value for each device attribute further comprises:
calculating for each value level cluster, by the device clustering module, a similarity index between the device and all devices in the value level cluster; and
assigning, by the device clustering module, the device to the value level cluster with the highest average similarity index.
6. The method of claim 5 wherein calculating for each value level cluster a similarity index between the device and all devices in the value level cluster further comprises:
for each value level cluster:
calculating, by the device clustering module, the similarity index between the device and each of the devices in the value level cluster;
summing, by the device clustering module, the similarity index between the device and each of the devices in the value level cluster, thereby producing a similarity index sum;
determining, by the device clustering module, the number of devices in the value level cluster; and
dividing, by the device clustering module, the similarity index sum by the number of devices in the value level cluster.
7. An apparatus for clustering devices in an Internet of Things (‘IoT’), the apparatus comprising a computer processor, a computer memory operatively coupled to the computer processor, the computer memory having disposed within it computer program instructions that, when executed by the computer processor, cause the apparatus to carry out the steps of:
receiving, by a device clustering module, a characteristic set for a device, wherein the characteristic set specifies one or more device attributes and an attribute value for each device attribute;
clustering, by the device clustering module, the device into an attribute level cluster based on the one or more device attributes specified in the characteristic set for the device; and
clustering, by the device clustering module, the device into a value level cluster based on the attribute value for each device attribute, wherein the value level cluster is a subset of the attribute level cluster.
8. The apparatus of claim 7 wherein the characteristic set is specified as an extensible markup language (‘XML’) structured document.
9. The apparatus of claim 7 wherein clustering the device into an attribute level cluster based on the one or more device attributes specified in the characteristic set for the device further comprises:
calculating, by the device clustering module, a commonality index between the device and a cluster representative for each attribute level cluster, wherein the cluster representative specifies a union of all attributes of each device in the attribute level cluster; and
assigning, by the device clustering module, the device to the attribute level cluster with the highest commonality index.
10. The apparatus of claim 9 wherein calculating a commonality index between the device and a cluster representative for each attribute level cluster further comprises:
determining, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative;
determining, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are not attributes of the cluster representative; and
dividing, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative by the sum of the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative and the number of device attributes specified in the characteristic set for the device that are not attributes of the cluster representative.
11. The apparatus of claim 7 wherein clustering the device into a value level cluster based on the attribute value for each device attribute further comprises:
calculating for each value level cluster, by the device clustering module, a similarity index between the device and all devices in the value level cluster; and
assigning, by the device clustering module, the device to the value level cluster with the highest average similarity index.
12. The apparatus of claim 11 wherein calculating for each value level cluster a similarity index between the device and all devices in the value level cluster further comprises:
for each value level cluster:
calculating, by the device clustering module, the similarity index between the device and each of the devices in the value level cluster;
summing, by the device clustering module, the similarity index between the device and each of the devices in the value level cluster, thereby producing a similarity index sum;
determining, by the device clustering module, the number of devices in the value level cluster; and
dividing, by the device clustering module, the similarity index sum by the number of devices in the value level cluster.
13. A computer program product for clustering devices in an Internet of Things (‘IoT’), the computer program product disposed upon a computer readable storage medium, wherein the computer readable storage medium is not a signal, the computer program product comprising computer program instructions that, when executed, cause a computer to carry out the steps of:
receiving, by a device clustering module, a characteristic set for a device, wherein the characteristic set specifies one or more device attributes and an attribute value for each device attribute;
clustering, by the device clustering module, the device into an attribute level cluster based on the one or more device attributes specified in the characteristic set for the device; and
clustering, by the device clustering module, the device into a value level cluster based on the attribute value for each device attribute, wherein the value level cluster is a subset of the attribute level cluster.
14. The computer program product of claim 13 wherein the characteristic set is specified as an extensible markup language (‘XML’) structured document.
15. The computer program product of claim 13 wherein clustering the device into an attribute level cluster based on the one or more device attributes specified in the characteristic set for the device further comprises:
calculating, by the device clustering module, a commonality index between the device and a cluster representative for each attribute level cluster, wherein the cluster representative specifies a union of all attributes of each device in the attribute level cluster; and
assigning, by the device clustering module, the device to the attribute level cluster with the highest commonality index.
16. The computer program product of claim 15 wherein calculating a commonality index between the device and a cluster representative for each attribute level cluster further comprises:
determining, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative;
determining, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are not attributes of the cluster representative; and
dividing, by the device clustering module, the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative by the sum of the number of device attributes specified in the characteristic set for the device that are also attributes of the cluster representative and the number of device attributes specified in the characteristic set for the device that are not attributes of the cluster representative.
17. The computer program product of claim 13 wherein clustering the device into a value level cluster based on the attribute value for each device attribute further comprises:
calculating for each value level cluster, by the device clustering module, a similarity index between the device and all devices in the value level cluster; and
assigning, by the device clustering module, the device to the value level cluster with the highest average similarity index.
18. The computer program product of claim 17 wherein calculating for each value level cluster a similarity index between the device and all devices in the value level cluster further comprises:
for each value level cluster:
calculating, by the device clustering module, the similarity index between the device and each of the devices in the value level cluster;
summing, by the device clustering module, the similarity index between the device and each of the devices in the value level cluster, thereby producing a similarity index sum;
determining, by the device clustering module, the number of devices in the value level cluster; and
dividing, by the device clustering module, the similarity index sum by the number of devices in the value level cluster.