Mechanism for synchronising devices, system and method

• US 10,055,397 B2
• Filed: 05/14/2013
• Issued: 08/21/2018
• Est. Priority Date: 05/14/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Touchtype Limited (Microsoft Corporation)
• Inventors: Bell, Michael, Freeman, Joe, Hategan, Emanuel George, Medlock, Benjamin
• Primary Examiner(s): WOZNIAK, JAMES S
• Application Number:
  US14/401,500
• Publication Number:
  US 20150134326A1
• Time in Patent Office: 1,925 Days
• Field of Search: 704 10, 704 1, 704 9, 715261
• CPC Class Codes: G06F 3/0237 using prediction or retrieval techniques
• CPC Class Codes: G06F 40/194 Calculation of difference between files
• CPC Class Codes: G06F 40/237 Lexical tools
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/274 Converting codes to words; Guess-ahead of partial word inputs

Abstract

There is provided a mechanism for synchronizing a plurality of dynamic language models residing in a plurality of devices associated with a single user, each device comprising a dynamic language model. The mechanism is configured to: receive text data representing text that has been input by a user into one or more of the plurality of devices; train at least one language model on the text data; and provide the at least one language model for synchronizing the devices. There is also provided a system comprising the mechanism and a plurality of devices, and a method for synchronizing a plurality of dynamic language models residing in a plurality of devices associated with a single user.

Claims

1. A system comprising:

a processor;

memory storing instructions that, when executed by the processor, configure the processor to;

receive text data representing text that has been input by a user into one or more of a plurality of devices associated with the user;

incorporate the text data into one or more language models by accumulating frequencies of occurrence for sequences of the text data, wherein the frequencies are for occurrences in respective ones of the language models; and

provide the one or more language models to one or more of the plurality of devices, the one or more language models being combined to merge the frequencies for occurrences in respective ones of the language models at each of the devices;

wherein, in response to deployment of the one or more language models among the plurality of devices, the one or more language models enable synchronized text prediction for text inputs among the plurality of devices based on the merged frequencies for occurrences.

2. The system of claim 1 wherein the text data comprises text and incorporating the text data into one or more language models comprises training the at least one language model on the text.

3. The system of claim 1 wherein the text data comprises a device delta language model trained on the text input by the user and incorporating the text data into one or more language models comprises merging the device delta language model into the one or more language models.

4. The system of claim 1, wherein incorporating the text data into one or more language models comprises generating at least one language model from the text data or incorporating the text data into at least one existing language model.

5. The system of claim 1, wherein incorporating the text data into one or more language models comprises incorporating the text data into a single language model to generate a cumulative language model.

6. The system of claim 5, wherein the text data represents text entered into any of the plurality of devices since a previous synchronisation of the devices and incorporating the text data into one or more language models comprises incorporating the text data into the cumulative language model.

7. The system of claim 5 further comprising:

a plurality of devices, each device comprising a dynamic language model;

wherein each of the plurality of devices is configured to transmit to the processor text data representing text entered into that device, and wherein each device is configured to receive the cumulative language model and merge the cumulative language model into its dynamic language model before transmitting to the processor the text data entered into that device.

8. The system of claim 7, wherein the one or more language models comprises a data structure associating sequences of terms with a frequency of occurrence for each sequence and the device is configured to merge a first language model with a second language model by:

adding the frequencies of occurrence for sequences in the data structure of he second language model to the frequencies of occurrence for corresponding sequences in the data structure of the first language model; and

inserting a new sequence and its corresponding frequency of occurrence into the data structure of the first language model, if that sequence is in the data structure of the second language model but not in the data structure of the first language model.

9. The system of claim 1, wherein incorporating the text data into one or more language models comprises generating a delta language model for each of the plurality of devices using the text data received from the plurality of devices except the device associated with the delta language model.

10. The system of claim 9 further comprising:

a plurality of devices, each device comprising a dynamic language model;

wherein the system is configured to synchronise with a single device of the plurality of devices at a given time; and

wherein each of the plurality of devices is configured to transmit to the processor text data representing text entered into that device, and wherein each device is configured to receive the delta language model associated with that device and merge the delta language model into its dynamic language model.

11. The system of claim 1, wherein the system is configured to synchronise with a single device of the plurality of devices at a given time.

12. The system of claim 1 further comprising:

a plurality of devices, each device comprising a dynamic language model;

wherein each of the plurality of devices is configured to transmit to the processor text data representing text that has been entered into that device.

13. The system of claim 12, wherein each dynamic language model of each device is configured to generate at least one text prediction based on text input into the device and wherein, once synchronised, the dynamic language models of the plurality of devices are capable of generating the same at least one text prediction when provided with the same text input.

14. The system of claim 12, wherein the system comprises a server and each of the plurality of devices is configured to download the one or more language models from the server and upload the text data onto the server.

15. The system of claim 1, wherein the one or more language models comprises a data structure associating sequences of terms with a frequency of occurrence for each sequence.

16. A method for synchronising a plurality of dynamic language models residing in a plurality of devices associated with a single user, each device comprising a dynamic language model, wherein the method comprises:

receiving, at a mechanism for synchronising, text data representing text that has been input by a user into one or more of the plurality of devices;

incorporating, with the mechanism, the text data into one or more language models by accumulating frequencies of occurrence for sequences of the text data, wherein the frequencies are for occurrences from each of the language models; and

providing the one or more language models, the one or more language models being combined to merge the frequencies for occurrences in respective ones of the language models at each of the devices;

wherein, in response to deployment of the one or more language models among the plurality of devices, the one or more language models enable synchronized text prediction for text inputs among the plurality of devices based on the merged frequencies for occurrences.

17. The method of claim 16, wherein the text data comprises text and incorporating the text data into one or more language models comprises training the at least one language model on the text.

18. The method of claim 16, wherein the text data comprises a device delta language model trained on the text input by the user and incorporating the text data into one or more language models comprises merging the device delta language model into the one or more language models.

19. The method of claim 18, wherein the one or more language models comprises a data structure associating sequences of terms with a frequency of occurrence for each sequence, and merging a first language model with a second language model comprises:

adding the frequencies of occurrence for sequences in the data structure of the second language model to the frequencies of occurrence for corresponding sequences in the data structure of the first language model; and

inserting a new sequence and its corresponding frequency of occurrence into the data structure of the first language model, if that sequence is in the data structure of the second language model but not in the data structure of the first language model.

20. The method of claim 16, wherein incorporating the text data into at least one language model comprises generating one or more language models from the text data or incorporating the text data into at least one existing language model.

21. The method of claim 16, wherein incorporating the text data into one or more language models comprises incorporating the text data into a single language model to generate a cumulative language model.

22. The method of claim 21, wherein the text data represents text entered into the plurality of devices since a previous synchronisation of the plurality of devices and incorporating the text data into one or more language models comprises incorporating the text data into the cumulative language model.

23. The method of claim 21, further comprising each device receiving the cumulative language model and merging the cumulative language model into the dynamic language model associated with that device prior to transmitting to the mechanism text data representing text entered into that device.

24. The method of claim 16, wherein each of the plurality of devices is synchronised individually with the mechanism.

25. The method of claim 16, wherein incorporating the text data into one or more language models comprises generating a delta language model for each of the plurality of devices using the text data received by the mechanism from the plurality of devices except the device associated with the delta language model.

26. The method of claim 25, wherein each of the plurality of devices is synchronised individually with the mechanism and wherein the text data incorporated into the delta language models comprises the text data received by the mechanism during a synchronisation and/or subsequent synchronisation of the plurality of devices except the device associated with the delta language model.

27. The method of claim 25, further comprising each device receiving the delta language model associated with that device and merging the delta language model into the dynamic language model of that device.

28. A non-transient computer readable medium containing program instructions which, when executed by a processor, configure the processor to carry out the method of claim 16.