Abstract: Embodiments relate to learning and determining user interface (UI) compliance with design guidelines. Those guidelines may enforce operability with requirements of an external UI platform. Embodiments receive as an input, a UI comprising a plurality of elements. An image of the UI is rendered, and a plurality of elements of the UI are identified from the image according to a recursive procedure. The plurality of elements are stored in a first logic tree exhibiting a first hierarchical structure having the plurality of elements as leaf nodes. The first logic tree is compared to a second logic tree exhibiting a second hierarchical structure, that is labeled with semantic metadata and stored within a knowledge base of a database. Input for the comparison may comprise the tree and/or additional meta information. Based upon the comparing, an output is generated indicating a similarity between the first logic tree and the second logic tree.

Abstract: A partitioned virtual space supports logical networking of IoT devices. Agents of the devices are assigned to interest-based cells in a virtual space, and can travel among the cells. Within the cells, pairs of devices are tested for similarity, based on device profiles, and for detected affinity. Agents of devices having affinity are connected in a logical network. Some attributes can be based on a personality model and can reflect the personality of a user or other principal associated with a device. Such attributes can influence requests for affinity testing, calculation of similarity, and further behavioral effects incorporated in affinity determination. Evaluation of recommendations can lead to updating of similarity scores or changes in affinity determination. Disclosed embodiments provide scalable, distributed, autonomous, and unsupervised device-to-device connectivity, free of prior constraints. Embodiments can be implemented in the cloud, with privacy protection.

Abstract: Attributes are applied to Internet-of-Things (IoT) devices to establish high quality connections between the devices. Agents of the devices are assigned to interest-based cells in a virtual space, and can travel among the cells. Within the cells, pairs of devices are tested for similarity, based on device profiles, and for detected affinity. Devices having affinity are connected and form a logical network of IoT devices. Some attributes can be based on a personality model and can reflect the personality of a user or other principal associated with a device. The user or principal attributes can influence requests for affinity testing, calculation of similarity, and further behavioral effects incorporated in affinity determination. Disclosed embodiments provide scalable, distributed, autonomous, and unsupervised device-to-device connectivity, free of prior constraints. Associated infrastructure, simulations, performance metrics, and variations are disclosed.

Abstract: A partitioned virtual space supports logical networking of IoT devices. Agents of the devices are assigned to interest-based cells in a virtual space, and can travel among the cells. Within the cells, pairs of devices are tested for similarity, based on device profiles, and for detected affinity. Agents of devices having affinity are connected in a logical network. Some attributes can be based on a personality model and can reflect the personality of a user or other principal associated with a device. Such attributes can influence requests for affinity testing, calculation of similarity, and further behavioral effects incorporated in affinity determination. Evaluation of recommendations can lead to updating of similarity scores or changes in affinity determination. Disclosed embodiments provide scalable, distributed, autonomous, and unsupervised device-to-device connectivity, free of prior constraints. Embodiments can be implemented in the cloud, with privacy protection.

Abstract: Attributes are applied to Internet-of-Things (IoT) devices to establish high quality connections between the devices. Agents of the devices are assigned to interest-based cells in a virtual space, and can travel among the cells. Within the cells, pairs of devices are tested for similarity, based on device profiles, and for detected affinity. Devices having affinity are connected and form a logical network of IoT devices. Some attributes can be based on a personality model and can reflect the personality of a user or other principal associated with a device. The user or principal attributes can influence requests for affinity testing, calculation of similarity, and further behavioral effects incorporated in affinity determination. Disclosed embodiments provide scalable, distributed, autonomous, and unsupervised device-to-device connectivity, free of prior constraints. Associated infrastructure, simulations, performance metrics, and variations are disclosed.

Abstract: Embodiments relate to learning and determining user interface (UI) compliance with design guidelines. Those guidelines may enforce operability with requirements of an external UI platform. Embodiments receive as an input, a UI comprising a plurality of elements. An image of the UI is rendered, and a plurality of elements of the UI are identified from the image according to a recursive procedure. The plurality of elements are stored in a first logic tree exhibiting a first hierarchical structure having the plurality of elements as leaf nodes. The first logic tree is compared to a second logic tree exhibiting a second hierarchical structure, that is labeled with semantic metadata and stored within a knowledge base of a database. Input for the comparison may comprise the tree and/or additional meta information. Based upon the comparing, an output is generated indicating a similarity between the first logic tree and the second logic tree.