Generation of context-aware, personalized challenges in computer games

1. A computer implemented method comprising:

receiving input data, the input data comprising data relating to a user of a computer game;

generating, based on the input data, one or more candidate challenges for the computer game, comprising:

determining one or more search constraints based on the input data; and

searching among one or more ranges of challenge parameters based on the one or more search constraints to generate the one or more candidate challenges;

determining, using a machine-learned model, whether each of the one or more candidate challenges satisfies a threshold condition, wherein the threshold condition is based on a target challenge difficultly;

in response to a positive determination, outputting the one or more candidate challenges that satisfy the threshold condition for use in the computer game by the user.

2. The computer implemented method of claim 1 , wherein the method further comprises:

in response to a negative determination that one or more candidate challenges satisfies the threshold condition:

updating the one or more search constraints based on output of the machine-learnedmodel;

re-searching among the one or more ranges of challenge parameters based on the one or more updated constraints to generate one or more further candidate challenges;

determining, using the machine-learned model, whether the one or more further candidate challenges satisfies the threshold condition; and

in response to a positive determination, outputting the one or more further candidate challenges that satisfy the threshold condition for use in the computer game by the user.

3. The computer implemented method of claim 1 , wherein determining one or more search constraints based on the input data comprises processing the input data using a game-specific logic.

4. The computer implemented method of claim 1 , wherein the one or more search constraints comprises: a cost-reward metric; a consistency condition for preventing mutually exclusive challenge parameters in each of the one or more candidate challenges; and/or game-specific in-game logic constraints.

5. The computer implemented method of claim 1 , wherein the input data comprises metric data generated from in-game data associated with the user of the computer game.

6. The computer implemented method of claim 5 , wherein the input data further comprises one or more of: in-game economic data; designer constraints; and/or real-world data extracted from a news feed using natural language processing.

7. The computer implemented method of claim 1 , wherein determining, using the machine-learned model, whether each of the one or more candidate challenges satisfies the threshold condition comprises:

generating, using the machine-learned model, a score for each of the one or more candidate challenges, the score indicative of a difficultly of the one or more candidate challenges; and

comparing the score to a threshold score associated with the user.

8. Apparatus comprising one or more processors and a memory, the memory comprising instructions that, when executed by the one or more processors, cause the apparatus to perform operations comprising:

receiving input data, the input data comprising data relating to a user of a computer game;

generating, based on the input data, one or more candidate challenges for the computer game, comprising:

determining one or more search constraints based on the input data; and

searching among one or more ranges of challenge parameters based on the one or more search constraints to generate the one or more candidate challenges;

determining, using a machine-learned model, whether each of the one or more candidate challenges satisfies a threshold condition, wherein the threshold condition is based on a target challenge difficultly;

in response to a positive determination, outputting the one or more candidate challenges that satisfy the threshold condition for use in the computer game by the user.

9. The apparatus of claim 8 , wherein the operations further comprise:

in response to a negative determination that the one or more candidate challenges satisfies the threshold condition:

updating the one or more search constraints based on output of the machine-learnedmodel;

re-searching among the one or more ranges of challenge parameters based on the one or more updated constraints to generate one or more further candidate challenges;

determining, using the machine-learned model, whether the one or more further candidate challenges satisfies the threshold condition; and

in response to a positive determination, outputting the one or more further candidate challenges that satisfy the threshold condition for use in the computer game by the user.

10. The apparatus of claim 8 , wherein determining one or more search constraints based on the input data comprises processing the input data using a game-specific logic.

11. The apparatus of claim 8 , wherein the one or more search constraints comprises: a cost-reward metric; a consistency condition for preventing mutually exclusive challenge parameters in each of the one or more candidate challenges; and/or game-specific in-game logic constraints.

12. The apparatus of claim 8 , wherein the input data comprises metric data generated from in-game data associated with the user of the computer game.

13. The apparatus of claim 12 , wherein the input data further comprises one or more of: in-game economic data; designer constraints; and/or real-world data extracted from a news feed using natural language processing.

14. The apparatus of claim 8 , wherein determining, using the machine-learned model, whether each of the one or more candidate challenges satisfies the threshold condition comprises:

generating, using the machine-learned model, a score for each of the one or more candidate challenges, the score indicative of a difficultly of the one or more candidate challenge; and

comparing the score to a threshold score associated with the user.

15. A computer implemented method comprising:

receiving player data from a plurality of players relating to a plurality of in-game challenges;

determining a proxy measure of difficulty for each of the plurality of in-game challenges based on the received player data;

processing each of the plurality of in-game challenges using one or more parametrised models to generate a prospective difficulty score for each of the plurality of in-game challenges;

comparing each of the prospective difficulty scores to a corresponding proxy measure of difficulty for the in-game challenge; and

updating parameters of the one or more parametrised models based on the comparisons to generate a machine-learned model,

wherein:

the plurality of players is divided into a plurality of player types in dependence on the received player data;

the proxy measure of difficulty for each of the plurality of in game-challenges comprises a plurality of difficulty measures, each corresponding to a player type; and

the one or more parameterised models comprises a plurality of parameterised models, each corresponding to one of the player types and updated based on data from the corresponding player type.

16. The computer implemented method of claim 15 , wherein the proxy measure of difficulty is based on one or more of: a number of players completing the challenge; a time taken to complete the challenge; a fraction of players that start the challenge and go on to complete the challenge; and/or a type of player who completes the challenge.

17. The computer implemented method of claim 15 , further comprising:

testing the machine-learned model on a test dataset comprising a plurality of in-game challenges; and

updating the proxy measure of difficulty based on the testing.