Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020136217 - METHOD AND SYSTEM FOR AUTOMATIC OPTIMIZATION OF USER'S BEHAVIOURAL CHANGES

Note: Text based on automatic Optical Character Recognition processes. Please use the PDF version for legal matters

[ EN ]

CLAIMS

1. A computer-implemented method for optimizing behavioural changes of a user, characterized by comprising the following steps:

setting a behavioural change goal for the user,

for a specific time instant, t,

adquiring static user characteristics (101) to generate a user characteristics vector (110),

adquiring dynamic user characteristics which define a current user state (102) to generate a user current state vector (120), querying (103) an artificial intelligence system (400) using the generated user characteristics vector (110) and user current state vector (120) as input to generate as output (104) a behavioural change tool vector (130) for the specific time instant, t, the output (104) being delivered to the user who makes one or more selections to change behaviour based on the output (104);

mapping the user characteristics vector (110), user current state vector (120) and the behavioural change tool vector (130) generated for the specific time instant, t, to a next current state vector (120) generated for a next time instant, t+1 ;

repeating the querying (103) to the artificial intelligence system (400) using as input the three vectors, user characteristics vector, user current state vector and behavioural change tool vector, updated over time, until the mapping indicates that the behavioural change goal is achieved; the selections taken by the user defining an optimal path for behavioural changes.

2. The method according to claim 1 , wherein the generation as output (104) of the behavioural change tool vector (130) by the artificial intelligence system (400) comprises reinforcement learning, wherein:

a subset of elements from the current state vector (120) generated for the specific time instant, t, represents a state space;

a subset of elements from the current state vector (120) and the user characteristics vector (110), both generated for the specific time instant, t, represents a context; and

a subset of elements in the next current state vector (121) generated for the next time instant, t+1 , represents a reward.

3. The method according to any preceeding claim, wherein the behavioural change tool vector (130) comprises a trigger for the user to take a selection from a set of actions, a content which is the set of actions, and an end which is a final action identifying the set behavioural change goal.

4. The method according to any preceeding claim, wherein the user current state vector (120) comprises three sub-vectors:

- a current behaviours sub-vector containing data referred to patterns of activities of daily living and of the behavioural change goal;

- a current dynamic personal characteristics sub-vector containing data referred to momentary cognitive processing;

- an emotional status sub-vector of emotional status containing data referred to assessment indices of user’s emotional and well-being.

5. The method according to any preceeding claim, wherein the user current state vector (120) is generated with dynamic user characteristics acquired by using internal or external services carried out by user’s terminal devices.

6. The method according to any preceeding claim, wherein the user current state vector (120) is generated with dynamic user characteristics acquired by ambient sensors.

7. The method according to any preceeding claim, wherein the user current state vector (120) is generated with dynamic user characteristics acquired by direct manual inputs from the user.

8. The method according to any preceeding claim, wherein the user characteristics vector (110) is generated with static user characteristics (101) comprising:

demographic information (210),

personal traits (220),

preferences (230) including a wide range of personal descriptors,

biological characteristics (240), and

positive and negative conceptual associations (250), defining object associations with respect to the behavioural change goal.

9. The method according to any preceeding claim, wherein the user characteristics vector (110) is generated with static user characteristics (101) adquired by using external services carried out by user’s terminal devices.

10. The method according to any preceeding claim, wherein the user characteristics vector (110) is generated with static user characteristics (101) adquired by external databases.

11. The method according to any preceeding claim, wherein the user characteristics vector (110) is generated with static user characteristics (101) adquired by ambient sensors.

12. The method according to any preceeding claim, wherein the user characteristics vector (110) is generated with static user characteristics (101) adquired by direct manual inputs from the user.

13. An artificial intelligence system (400) for optimizing behavioural changes of a user characterized by comprising processing means implementing artificial intelligence (410) and a database (420) on which the artificial intelligence (410) is built, and the artificial intelligence (410) being configured to, for a behavioural change goal set for the user:

generate a user characteristics vector (110) for a specific time instant, t, using adquired static user characteristics (101);

generate a user current state vector (120) for a specific time instant, t, using adquired dynamic user characteristics which define a current user state (102); generate a behavioural change tool vector (130) for the specific time instant, t, and deliver the behavioural change tool vector (130) as an output (104) from the artificial intelligence system (400) to the user, who makes one or more selections to change behaviour based on the behavioural change tool vector

(130); the behavioural change tool vector (130) being generated using the generated user characteristics vector (110) and user current state vector (120) as inputs to the artificial intelligence system (400) in a query (103);

map the user characteristics vector (110), user current state vector (120) and the behavioural change tool vector (130) generated for the specific time instant, t, to a next current state vector (120) generated for a next time instant, t+1 ;

repeat the query (103) using as input the three vectors, user characteristics vector, user current state vector and behavioural change tool vector, updated over time, until the mapping indicates that the behavioural change goal is achieved; the selections taken by the user defining an optimal path for behavioural changes.

14. The system (400) according to claim 13, wherein the artificial intelligence (410) is configured to generate the behavioural change tool vector (130) by reinforcement learning, wherein:

a subset of elements from the current state vector (120) generated for the specific time instant, t, represents a state space;

a subset of elements from the current state vector (120) and the user characteristics vector (110), both generated for the specific time instant, t, represents a context; and

a subset of elements in the next current state vector (121) generated for the next time instant, t+1 , represents a reward.

15. The system (400) according to any of claims 13-14, wherein the artificial intelligence (410) is configured to generate the behavioural change tool vector (130) comprising a trigger for the user to take a selection from a set of actions, a content which is the set of actions, and an end which is a final action identifying the set behavioural change goal.

16. The system (400) according to any of claims 13-15, wherein the artificial intelligence (410) is configured to generate the user current state vector (120) comprising three sub-vectors:

- a current behaviours sub-vector containing data referred to patterns of activities of daily living and of the behavioural change goal;

- a current dynamic personal characteristics sub-vector containing data referred to momentary cognitive processing;

- an emotional status sub-vector of emotional status containing data referred to assessment indices of user’s emotional and well-being.

17. The system (400) according to any of claims 13-16, wherein the artificial intelligence (410) is configured to generate the user current state vector (120) with dynamic user characteristics acquired by using internal or external services carried out by user’s terminal devices.

18. The system (400) according to any of claims 13-17, wherein the artificial intelligence (410) is configured to generate the user current state vector (120) with dynamic user characteristics acquired by ambient sensors.

19. The system (400) according to any of claims 13-18, wherein the artificial intelligence (410) is configured to generate the user current state vector (120) with dynamic user characteristics acquired by direct manual inputs from the user.

20. The system (400) according to any of claims 13-19, wherein the artificial intelligence (410) is configured to generate the user characteristics vector (110) with static user characteristics (101) comprising:

demographic information (210),

personal traits (220),

preferences (230) including a wide range of personal descriptors,

biological characteristics (240), and

positive and negative conceptual associations (250), defining object associations with respect to the behavioural change goal.

21. The system (400) according to any of claims 13-20, wherein the artificial intelligence (410) is configured to generate the user characteristics vector (110) with static user characteristics (101) adquired by using external services carried out by user’s terminal devices.

22. The system (400) according to any of claims 13-21 , wherein the artificial intelligence (410) is configured to generate the user characteristics vector (110) with static user characteristics (101) adquired by external databases.

23. The system (400) according to any of claims 13-22, wherein the artificial intelligence (410) is configured to generate the user characteristics vector (110) with static user characteristics (101) adquired by ambient sensors.

24. The system (400) according to any of claims 13-20, wherein the artificial intelligence (410) is configured to generate the user characteristics vector (110) with static user characteristics (101) adquired by direct manual inputs from the user.