Generating digital event recommendation sequences utilizing a dynamic user preference interface

Patent metadata

Field	Value
Publication number	US-11946753-B2
Application number	US-202117364480-A
Country	US
Kind code	B2
Filing date	Jun 30, 2021
Priority date	Jul 27, 2018
Publication date	Apr 2, 2024
Grant date	Apr 2, 2024

Abstract

The present disclosure relates to generating and modifying recommended event sequences utilizing a dynamic user preference interface. For example, in one or more embodiments, the system generates a recommended event sequence using a recommendation model trained based on a plurality of historical event sequences. The system then provides, for display via a client device, the recommendation, a plurality of interactive elements for entry of user preferences, and a visual representation of historical event sequences. Upon detecting input of user preferences, the system can modify a reward function of the recommendation model and provide a modified recommended event sequence together with the plurality of interactive elements. In one or more embodiments, as a user enters user preferences, the system additionally modifies the visual representation to display subsets of the plurality of historical event sequences corresponding to the preferences.

First claim

What is claimed is:

1. A computer-implemented method comprising:
generating, for a user of a client device, a popular event sequence based on a frequency of use of event sequences by a plurality of prior users represented in a plurality of historical event sequences;
generating a recommended event sequence by using a recommendation machine learning model to select an event sequence for recommendation based on a reward function having a plurality of parameters that were learned during multiple training iterations to provide expected values of recommendations, the recommended event sequence corresponding to a general recommendation provided to client devices users as a default;
receiving, from the client device, one or more user preferences with respect to one or more events by receiving at least one user interaction with one or more interactive elements corresponding to the one or more events via a graphical user interface of the client device;
generating a modified recommended event sequence using the recommendation machine learning model by modifying the reward function to include a weighting factor that modifies the plurality of parameters of the reward function via one or more preference weights that represent the one or more user preferences to modify how the recommendation machine learning model selects the event sequence for recommendation without retraining the recommendation machine learning model; and
providing, for simultaneous display within the graphical user interface on the client device, the recommended event sequence, the modified recommended event sequence, and the popular event sequence.

2. The computer-implemented method of claim 1 , further comprising:
receiving, from the client device, one or more user constraints that place a limitation on at least one event by receiving one or more user interactions with one or more additional interactive elements; and
modifying the reward function of the recommendation machine learning model based on the one or more user constraints.

3. The computer-implemented method of claim 1 , further comprising:
determining a first expected value score corresponding to the modified recommended event sequence using the modified reward function; and
providing, for display within the graphical user interface on the client device, the first expected value score in conjunction with the modified recommended event sequence.

4. The computer-implemented method of claim 3 , further comprising:
determining, using the modified reward function, a second expected value score corresponding to the popular event sequence;
determining, using the modified reward function, a third expected value score corresponding to the recommended event sequence; and
providing, for display within the graphical user interface on the client device, the second expected value score in conjunction with the popular event sequence and the third expected value score in conjunction with the recommended event sequence.

5. The computer-implemented method of claim 1 , wherein modifying the reward function of the recommendation machine learning model to introduce the weighting factor that modifies the plurality of parameters of the reward function via the one or more preference weights that represent the one or more user preferences comprises providing a preference weight to each of the plurality of parameters of the reward function where each preference weight represents a corresponding user preference.

6. The computer-implemented method of claim 1 , further comprising:
providing, for display within the graphical user interface on the client device, a first set of visual indicators corresponding to events of the recommended event sequence;
providing, for display within the graphical user interface on the client device, a second set of visual indicators corresponding to events of the modified recommended event sequence; and
providing, for display within the graphical user interface on the client device, a third set of visual indicators corresponding to events of the popular event sequence.

7. The computer-implemented method of claim 6 ,
further comprising determining that a first event from the events of the modified recommended event sequence is more frequently used by the plurality of prior users than a second event from the events of the modified recommended event sequence,
wherein providing the third set of visual indicators corresponding to the events of the modified recommended event sequence comprises providing a first visual indicator corresponding to the first event having a larger size than a second visual indicator corresponding to the second event based on determining that the first event is more frequently used than the second event.

8. The computer-implemented method of claim 1 , wherein:
receiving the one or more user preferences comprises receiving at least one of a preferred point of interest category or a preferred point of interest; and
generating the modified recommended event sequence based on the one or more user preferences comprises generating the modified recommended event sequence to include at least one of a point of interest associated with the preferred point of interest category or to include the preferred point of interest.

9. The computer-implemented method of claim 1 , further comprising:
receiving, from an additional client device, one or more additional user preferences with respect to the one or more events;
generating, utilizing the recommendation machine learning model, an additional modified recommended event sequence based on the one or more additional user preferences; and
providing, for display within a graphical user interface on the additional client device, the recommended event sequence, the additional modified recommended event sequence, and the popular event sequence.

10. A non-transitory computer-readable medium storing instructions thereon that, when executed by at least one processor, cause the at least one processor to perform operations comprising:
generating, for a user of a client device, a popular event sequence based on a frequency of use of event sequences by a plurality of users represented in a plurality of historical event sequences;
generating a recommended event sequence by using a recommendation machine learning model to select an event sequence for recommendation based on a reward function having a plurality of parameters that were learned during multiple training iterations to provide expected values of recommendations, the recommended event sequence corresponding to a general recommendation provided to client device users as a default;
providing, for display within a graphical user interface of the client device, the popular event sequence, the recommended event sequence, and a plurality of interactive elements for entry of user preferences events;
determining one or more user preferences with respect to one or more events based on receiving at least one user interaction with the plurality of interactive elements via the graphical user interface of the client device; and
in response to determining the one or more user preferences:
generating a modified recommended event sequence using the recommendation machine learning model by modifying the reward function to include a weighting factor that modifies that plurality of parameters of the reward function via one or more preference weights that represent the one or more user preferences to modify how the recommendation machine learning model selects the event sequence for recommendation without retraining the recommendation machine learning model; and
modifying the graphical user interface of the client device to simultaneously display the recommended event sequence, the modi

Assignees

• Adobe Inc

Inventors

• Du Fan
• Lee Sana Malik
• Theocharous Georgios
• Koh Eunyee

Classifications

• G01C21/343 Primary

  Calculating itineraries (travelling salesman problem G06Q10/04; optimisation of routes G06Q10/047) · CPC title

• G01C21/3476

  using point of interest [POI] information, e.g. a route passing visible POIs · CPC title

• G01C21/3484

  Personalized, e.g. from learned user behaviour or user-defined profiles · CPC title

• G06Q10/047 Primary

  Optimisation of routes or paths, e.g. travelling salesman problem · CPC title

• H04W4/021

  Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences · CPC title