Method of label transform for managing heterogeneous information

• US 10,565,242 B2
• Filed: 01/10/2017
• Issued: 02/18/2020
• Est. Priority Date: 01/10/2017
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Anglin, Debbie A., Liu, Su, Ly, Andrew, Yan, Shunguo
• Primary Examiner(s): Trujillo, James
• Assistant Examiner(s): Bogacki, Michal
• Application Number:
  US15/403,068
• Publication Number:
  US 20180196872A1
• Time in Patent Office: 1,134 Days
• CPC Class Codes: G06F 16/3346 using probabilistic model
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: H04L 67/12 specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
• CPC Class Codes: H04L 67/565 Conversion or adaptation of application format or content adding application control or application functional data H04L67/561
• CPC Class Codes: H04W 4/70 Services for machine-to-machine communication [M2M] or machine type communication [MTC]

Abstract

Techniques are disclosed herein for normalizing heterogeneous information provided in labels processed by Internet of Things (IoT) devices. A device (e.g., an IoT-enabled device) receives an item including label content. The label content includes attributes describing the item. The device determines that at least one of the attributes is not formatted for processing by the device. A request is generated to format the attributes for processing. The request includes the label content and a target language and a target encoding. The request is sent to a network service that translates the label content. The device receives the translated label content from the network service and processes the item according to the translated label content.

Claims

1. A method comprising:

receiving, by a device, an image of a label attached to a garment, wherein the label includes a plurality of care instructions for cleaning the garment;

upon the device determining that a first instruction of the plurality of instructions cannot be interpreted for processing by the device, evaluating with a local machine learning model on the device an item history for the garment to generate a first score for a first replacement care instruction based on patterns identified from the label using the local machine learning model, wherein the first score indicates a likelihood that the first instruction corresponds to the first replacement care instruction, and wherein the first historical care instruction is appropriately formatted for processing by the device;

upon determining that the first score does not exceed a predefined threshold, generating a request to translate the first instruction into an appropriate format for processing by the device, wherein the request includes the image and a target format;

sending the request to a network service for translating the first instruction;

receiving translated label content from the network service, wherein the translated label content comprises at least a translated first instruction; and

processing, by the device, the item according to the translated label content.

2. The method of claim 1, further comprising:

upon the device determining that a second instruction cannot be interpreted for processing by the device, evaluating with the local machine learning model on the device the item history for the garment to generate a second score for a second replacement instruction based on the patterns identified from the label using the local machine learning model, wherein the second score indicates a likelihood that the second instruction corresponds to the second replacement instruction, and wherein the second replacement instruction is appropriately formatted for processing by the device.

3. The method of claim 2, further comprising:

upon determining that the second score exceeds the predefined threshold, processing the garment according to the second replacement instruction.

4. The method of claim 1, wherein the device is an IoT-enabled washing machine.

5. The method of claim 1, wherein the label includes attributes for the garment that include one or more of an item type, item group, and item information.

6. The method of claim 1, wherein determining that the first instruction cannot be interpreted for processing by the device comprises:

determining that a language and an encoding associated with first instruction do not match the target format.

18. The method of claim 1, further comprising:

upon receiving the translated label content from the network service, determining an appropriate action relating to processing the garment;

outputting the appropriate action to a mobile device that is different from the device and the network service;

receiving input from the mobile device; and

processing the garment by the device according to the input from the mobile device.

7. A computer program product, comprising:

a computer-readable storage medium storing instructions, which, when executed on a processor, perform an operation, comprising;

receiving, by a device, an image of a label attached to a garment, wherein the label includes a plurality of care instructions for cleaning the garment;

upon the device determining that a first instruction of the plurality of instructions cannot be interpreted for processing by the device, evaluating with a local machine learning model on the device an item history for the garment to generate a first score for a first replacement care instruction based on patterns identified from the label using the local machine learning model, wherein the first score indicates a likelihood that the first instruction corresponds to the first replacement care instruction, and wherein the first historical care instruction is appropriately formatted for processing by the device;

upon determining that the first score does not exceed a predefined threshold, generating a request to translate the first instruction into an appropriate format for processing by the device, wherein the request includes the image and a target format;

sending the request to a network service for translating the first instruction;

receiving translated label content from the network service, wherein the translated label content comprises at least a translated first instruction; and

processing, by the device, the item according to the translated label content.

8. The computer program product of claim 7, wherein the operation further comprises:

upon the device determining that a second instruction cannot be interpreted for processing by the device, evaluating with the local machine learning model on the device the item history for the garment to generate a second score for a second replacement instruction based on the patterns identified from the label using the local machine learning model, wherein the second score indicates a likelihood that the second instruction corresponds to the second replacement instruction, and wherein the second replacement instruction is appropriately formatted for processing by the device.

9. The computer program product of claim 8, wherein the operation further comprises:

upon determining that the second score exceeds the predefined threshold, processing the garment according to the second replacement instruction.

10. The computer program product of claim 7, wherein the device is an IoT-enabled washing machine.

11. The computer program product of claim 7, wherein the label include attributes of the garment that include one or more of an item type, item group, and item information.

12. The computer program product of claim 7, wherein determining that the first instruction cannot be interpreted for processing by the device comprises:

determining that a language and an encoding associated with the first instruction do not match the target format.

19. The computer program product of claim 7, wherein the operation further comprises:

upon receiving the translated label content from the network service, determining an appropriate action relating to processing the garment;

outputting the appropriate action to a mobile device that is different from the device and the network service;

receiving input from the mobile device; and

processing the garment by the device according to the input from the mobile device.

13. A system, comprising:

a processor; and

a memory storing program code, which, when executed on the processor, performs an operation, comprising;

receiving, by a device, an image of a label attached to a garment, wherein the label includes a plurality of care instructions for cleaning the garment;

upon the device determining that a first instruction of the plurality of instructions cannot be interpreted for processing by the device, evaluating with a local machine learning model on the device an item history for the garment to generate a first score for a first replacement care instruction based on patterns identified from the label using the local machine learning model, wherein the first score indicates a likelihood that the first instruction corresponds to the first replacement care instruction, and wherein the first historical care instruction is appropriately formatted for processing by the device;

upon determining that the first score does not exceed a predefined threshold, generating a request to translate the first instruction into an appropriate format for processing by the device, wherein the request includes the image and a target format;

sending the request to a network service for translating the first instruction;

receiving translated label content from the network service, wherein the translated label content comprises at least a translated first instruction; and

processing, by the device, the item according to the translated label content.

14. The system of claim 13, wherein the operation further comprises:

upon the device determining that a second instruction cannot be interpreted for processing by the device, evaluating with the local machine learning model on the device the item history for the garment to generate a second score for a second replacement instruction based on the patterns identified from the label using the local machine learning model, wherein the second score indicates a likelihood that the second instruction corresponds to the second replacement instruction, and wherein the second replacement instruction is appropriately formatted for processing by the device.

15. The system of claim 14, wherein the operation further comprises:

upon determining that the second score exceeds the predefined threshold, processing the garment according to the second replacement instruction.

16. The system of claim 13, wherein the device is an IoT-enabled washing machine.

17. The system of claim 13, wherein the label includes attributes that describe the garment that include one or more of an item type, item group, and item information.

20. The system of claim 13, wherein the operation further comprises:

upon receiving the translated label content from the network service, determining an appropriate action relating to processing the garment;

outputting the appropriate action to a mobile device that is different from the device and the network service;

receiving input from the mobile device; and

processing the garment by the device according to the input from the mobile device.