Dynamic contextual data capture
AssigneeSap SE
InventorJackson Mathai
Life of this patent
Estimating the term…
Abstract
Disclosed herein are system, method, and computer program product embodiments for scanning and identifying a physical object. An individual may send the scanned object to a recognition engine. The recognition engine may determine the nature of the physical object using an object recognition model and a semantic meaning for any written text on the physical object using a natural language processing model. User-entered context may be displayed for and received from users related to the identified physical object. An appropriate set of actions to perform in response to the scan may be determined and subsequently performed on the mobile device.
Claims20
A computer-implemented method, comprising:
- scanning a physical object comprising written text in a scan window on a mobile device;
- determining, by a recognition engine, a setting applicable to the physical object using an object recognition model;
- deriving, by the recognition engine, a semantic meaning for the written text using a natural language processing model that considers the setting;
- selecting, by the recognition engine, a set of actions performable on the mobile device based on the semantic meaning and an artificial intelligence model;
- causing execution of an action in the set of actions on the mobile device;
- and
- updating the artificial intelligence model based on the action and a satisfaction indicator,
- wherein at least one of the scanning, determining, deriving, selecting, causing and updating are performed by one or more computers.
A system operable with a mobile device, the mobile device configured to scan a physical object comprising written text in a scan window, and execute an action in a set of actions, the system comprising:
- a memory, and
- at least one processor coupled to the memory and configured to:
- determine a setting applicable to the physical object using an object recognition model;
- derive a semantic meaning for the written text using a natural language processing model that considers the setting;
- select the set of actions based on the semantic meaning and an artificial intelligence model;
- and
- cause execution of the action in the set of actions on the mobile device;
- and
- update the artificial intelligence model based on the action and a satisfaction indicator.
A non-transitory computer-readable device having instructions stored thereon that, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:
- scanning a physical object comprising written text in a scan window on a mobile device;
- determining, by a recognition engine, a setting applicable to the physical object using an object recognition model;
- deriving, by the recognition engine, a semantic meaning for the written text using a natural language processing model that considers the setting;
- selecting, by the recognition engine, a set of actions performable on the mobile device based on the semantic meaning and an artificial intelligence model;
- causing execution of an action in the set of actions on the mobile device;
- and
- updating the artificial intelligence model based on the action and a satisfaction indicator.
Dependencies are inferred from the claim text.