WO2023125692A1 - Procédé de recommandation de service et appareil associé - Google Patents

Procédé de recommandation de service et appareil associé Download PDF

Info

Publication number
WO2023125692A1
WO2023125692A1 PCT/CN2022/142915 CN2022142915W WO2023125692A1 WO 2023125692 A1 WO2023125692 A1 WO 2023125692A1 CN 2022142915 W CN2022142915 W CN 2022142915W WO 2023125692 A1 WO2023125692 A1 WO 2023125692A1
Authority
WO
WIPO (PCT)
Prior art keywords
electronic device
user
service
service set
time
Prior art date
Application number
PCT/CN2022/142915
Other languages
English (en)
Chinese (zh)
Inventor
王云路
邹志国
方荣杰
李序旸
唐鹏程
刘成
卓晓燕
朱君浩
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2023125692A1 publication Critical patent/WO2023125692A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/14—Digital output to display device ; Cooperation and interconnection of the display device with other functional units
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/14—Digital output to display device ; Cooperation and interconnection of the display device with other functional units
    • G06F3/1423—Digital output to display device ; Cooperation and interconnection of the display device with other functional units controlling a plurality of local displays, e.g. CRT and flat panel display
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services

Definitions

  • terminals such as mobile phones, tablet computers, and smart wearable devices have been popularized.
  • the terminal can use various services provided by the application program or the system to realize corresponding functions.
  • viewing the booking information in application A requires the user to independently search for the corresponding application and open the relevant function interface of the application through multiple operations, which are relatively cumbersome operations. How to improve the convenience and accuracy of service recommendation is an urgent problem to be solved.
  • the method further includes: the first electronic device generates a second service set based on the personal data of the first user, the second service set includes multiple service functions, and the second service set corresponds to The multiple functional operations are different from the multiple functional operations corresponding to the first service set; when it is detected that the trigger condition of the second service set is met, the first electronic device executes the multiple functional operations corresponding to the second service set.
  • the first electronic device can automatically generate different service sets for the user, reflecting the intelligence of service recommendation.
  • the first electronic device can dynamically adjust the time for the first electronic device to display the first service set based on the time when the user uses the first service set and the time for displaying the first service set, so as to prevent the first electronic device from displaying the first service set. There is a large deviation between the time of the service set and the time when the user uses the first service set.
  • the displaying of the first recommendation card by the first electronic device at the third time specifically includes: the first electronic device detects that the first new sub-condition is executed, and the first electronic device The first recommendation card is displayed at a third time.
  • the method further includes: when the device displaying the first service set is the second electronic device, the first electronic device sends the identifier of the service function in the first service set to the second electronic device. On the electronic device, the identifier of the service function in the first service set is used by the second electronic device to display the first recommendation card.
  • the present application provides an electronic device, which is a first electronic device, characterized in that the first electronic device includes: one or more processors, one or more memories; one or more memories and one or more A plurality of processors are coupled, and one or more memories are used to store computer program codes, and the computer program codes include computer instructions, and one or more processors invoke computer instructions to make the first electronic device perform any of the above-mentioned aspects.
  • a service recommendation method provided in the implementation.
  • FIG. 35 exemplarily shows a schematic flowchart of a service recommendation method provided by an embodiment of the present application.
  • Application services are services provided by providers of applications installed on electronic devices. Different applications can provide the same service, for example, a taxi application can provide a taxi service, and a map application can also provide a navigation service. The same application can also provide multiple services with different functions, for example, the WeChat application can provide payment services, and the WeChat application can also provide call services, etc.
  • the electronic device 100 may record the time when the user triggers the electronic device 100 to execute the one or more services, and based on the actual execution time of the one or more services and the time when the electronic device 100 displays the one or more services, determine that the electronic device 100 100 whether the time for displaying one or more services is appropriate, if not, electronic device 100 will adjust the time for electronic device 100 to display one or more services based on preset rules, so that electronic device 100 displays one or more services
  • the difference between the time of the electronic device 100 and the time when the electronic device 100 actually executes the one or more services is less than the preset time, which improves the accuracy of the electronic device 100 in displaying the service card.
  • the principle of how to adjust the launch condition of the service is similar to the principle of how to adjust the trigger condition of the service. Specifically, how the electronic device 100 adjusts the triggering condition of the service may refer to the embodiments introduced in FIG. 21-FIG. 22 , and details will not be described here in this embodiment.
  • the gyro sensor 180B can be used to determine the motion posture of the electronic device 100 .
  • the angular velocity of the electronic device 100 around three axes may be determined by the gyro sensor 180B.
  • the acceleration sensor 180E can detect the acceleration of the electronic device 100 in various directions (generally three axes).
  • the distance sensor 180F is used to measure the distance.
  • Proximity light sensor 180G may include, for example, light emitting diodes (LEDs) and light detectors, such as photodiodes.
  • the light emitting diodes may be infrared light emitting diodes.
  • the electronic device 100 emits infrared light through the light emitting diode.
  • the server 200 can also calculate the probability of multiple services that the user will use at the same location type, and calculate the probability of multiple services that the same location type will use
  • the probabilities are sorted from largest to smallest. Exemplarily, when a user is at a place where the place type is food, the service with the highest probability of use is food service, followed by payment service, and finally entertainment service. When a user is in a place where the location type is leisure and entertainment, the service with the highest probability of use is entertainment service, followed by sports service, and finally transportation service. When a user is in a place where the location type is sports and fitness, the service with the highest probability of use is sports service, followed by entertainment service, and finally schedule service.
  • the electronic device 100 may also perform preprocessing on the user data. For example, standardize the processing of various types of data and classify them according to their importance, sensitivity, timeliness or other specifics, and establish a unified data storage mode. In this way, user data obtained through different methods and platforms can be stored in a unified standard, and complete and comprehensive data can be collected to facilitate subsequent construction of complete user attributes.
  • different types of user data can have different storage standards, for example, exercise health data and travel data can be stored by different standards.
  • the electronic device 100 may delete the information that the service is the takeaway service.
  • the electronic device 100 may also trigger the recommendation process in other ways, which is not limited in this embodiment of the present application.
  • Multiple approaches may include, but are not limited to, a scene-bound calibration service set, a historical click-through rate service list, a service interest list, a reminder service list, and the like.
  • the training samples can be obtained through multiple channels, which is not limited in the embodiment of the present application.
  • the input of the training samples can be the time, place and the state of the electronic device (such as Bluetooth open state, data connection state, earphone connection state, etc. etc.), the output of the training sample can be the service actually used by the user.
  • the electronic device 100 uses the training data 701 to iteratively train the preset model based on the preset model to obtain the trained model 702 .
  • Execution conditions may include, but are not limited to: the user's trigger operation on the recommended card, the user's voice command meeting the preset voice command, and the user's identity information meeting the preset identity information.
  • the electronic device 100 may perform multiple functional operations based on the execution relationship of multiple services in the service set displayed on the recommendation card. For example, functional operations corresponding to multiple services may be executed simultaneously, or multiple services may be executed in a certain sequence, and so on.
  • the electronic device 100 may trigger the electronic device 100 to execute the service set shown in the recommendation card based on any of the following methods.
  • the card 801 shows the smart home linkage service recommended by the electronic device 100 for user A in the scene of returning home.
  • the card 801 may be composed of text, pictures and other elements.
  • the electronic device 100 respectively sends opening instructions to the air purifier in the living room, the air conditioner in the living room, and the curtains in the living room. .
  • the electronic device 100 After the electronic device 100 obtains the service to be recommended in the home scene, the electronic device 100 will generate a recommendation card based on the service to be recommended in the home scene, and the recommendation card may be a card 802 as shown in FIG. 9C .
  • the electronic device 100 can automatically generate multiple services with different functions for the user at the same time, reducing user operations.
  • services can be divided into system services and third-party services, where third-party services are services provided by non-electronic device 100 manufacturers, and system services are services provided by electronic device 100 manufacturers. Services can also be divided according to functions, which is not limited in this embodiment of the present application.
  • the electronic device 100 can freely combine system services and third-party services to generate service sets in different combinations.
  • 9I-9K exemplarily show schematic views of the electronic device 100 displaying a recommendation card of a third-party service.
  • FIG. 9L exemplarily shows a schematic diagram of electronic device 100 simultaneously displaying recommendation cards of third-party services and system services.
  • the electronic device 100 displays a recommendation card, the probability of the service shown on the recommendation card is less than a preset value, and when the electronic device 100 detects that the user does not use the service set shown on the recommendation card, or the user manually closes the recommendation card, in this case, the electronic device 100 can update the probability estimation module, so that the probability that the probability estimation module estimates the above-mentioned service set that the user may use is further reduced, and avoid generating services that the user will not use. Recommendation card.
  • the electronic device 100 can learn the usage habits of the user on the electronic device 100, actively generate a service set that the user may use at different times and places, and generate recommendations based on the service set card. In this way, on the one hand, the user is not required to manually set the services to be used, which reduces user operations.
  • the electronic device 100 recommends a service set for the user based on the usage habits of the user. The more usage data of the user, the more accurate the service set recommended by the electronic device 100 is, and the service set recommended by the electronic device 100 is more in line with the behavior characteristics of the user.
  • the multiple service sets in each scene may be preset service sets in the electronic device 100 .
  • the electronic device 100 filters out a service set that the user may use from the preset services, and recommends the service set to the user.
  • the electronic device 100 may, in combination with the user's degree of attention to the electronic devices associated with the user, recommend to the user the services used by similar users whose types of devices are close to the ones they follow. Collection of services.
  • the server 200 Before the server 200 calculates the similarity of the service sets used by different users based on the similarity calculation model, the server 200 needs to train the similarity calculation model so that the similarity calculation model can accurately calculate the similarity of the service sets used by different users.
  • the training samples can be obtained through various channels, which is not limited in this embodiment of the present application.
  • the input of the training sample can be a set of services used by different users in the same scenario, and the output of the training sample can be the value of the similarity between different users.
  • the server 200 uses the training data to iteratively train the similarity calculation model based on the preset model to obtain a trained similarity calculation model.
  • Table 3 exemplifies the collection of services used by different users in the movie viewing scene, which are collected by the server 200 .
  • the set of services used by user A in the movie viewing scene includes: turning on the do-not-disturb mode of the mobile phone, and turning off the lights in the living room.
  • the set of services used by user B in the movie viewing scene includes: turning on the do-not-disturb mode on the mobile phone, turning off the lights in the living room, and closing the curtains in the living room.
  • the set of services used by user C in the movie viewing scene includes: turning off the do-not-disturb mode of the mobile phone, and turning on the lights in the living room.
  • the set of services used by user D in the movie viewing scene includes: turning on the do-not-disturb mode on the mobile phone, turning off the lights in the living room, and turning on the air conditioner in the living room to 26 degrees.
  • the server 200 calculates the similarity between the vectors to obtain the similarity of service sets used by different users in the same scene.
  • the server 200 can calculate the similarity between vectors by any of the following methods: cosine similarity between vectors, Euclidean distance between vectors, Manhattan distance between vectors, Pearson correlation between vectors Coefficient etc.
  • the server 200 may also calculate the similarity of service sets used by different users in the same scenario in other ways.
  • the electronic device 100 can recommend to user A a service set used by other users whose similarity with the service set used by user A ranks in the top N (for example, N is equal to 1).
  • the electronic device 100 will generate a recommendation card based on the service set used by user B and the service set used by user D, and display it on the desktop of the electronic device 100 .
  • FIG. 12 exemplarily shows a recommendation card 1200 generated by the electronic device 100 .
  • the recommendation card 1200 may include the history service collection card 1201 used by user A in the movie watching scene and the service collection card (namely card 1202 ) recommended by the electronic device 100 for user A to be used by similar users.
  • the card can be composed of text, pictures and other elements.
  • the services displayed in the card 1201 include turning on the do-not-disturb mode of the mobile phone and turning off the lights in the living room.
  • Card 1202 is a collection of services used by user B in the movie watching scene.
  • the services displayed in card 1202 include turning on the do-not-disturb mode of the mobile phone, turning off the lights in the living room, and closing the curtains in the living room.
  • the electronic device may score the degree of attention of the device associated with the user, and reorder the set of services used by similar users obtained above based on the user's degree of attention to the device associated with the user to obtain the final recommendation result .
  • the electronic device 100 can recommend to the user a set of services used by similar users that is close to the type of device that the user cares about, so that the set of services recommended by the electronic device 100 is more accurate.
  • the attention degree of the user to the device can be obtained based on the following information: the working state of the device associated with the user, and the state of the user detected by the device associated with the user.
  • the working state of the device associated with the user includes, but not limited to: whether the device associated with the user is turned on, and whether the device associated with the user is in a working state (for example, whether there is image or audio output).
  • the state of the user detected by the device associated with the user includes but is not limited to: whether the user is staring at the large screen or the mobile phone for a long time.
  • the electronic device 100 may allow the user to select a device that the user needs to pay attention to by means of a pop-up window.
  • the device selected by the user has an attention score of 1. For example, in a movie watching scene, if the user selects a large screen to focus on, the electronic device 100 obtains a score of 1 for the user's attention to the large screen. In the driving scene, if the user selects the car computer to be concerned about, the electronic device 100 obtains a score of 1 for the user's attention to the car computer.
  • the electronic device 100 needs to acquire multiple training samples.
  • the training samples can be obtained through various channels, which is not limited in this embodiment of the present application.
  • the training sample includes the input of the training sample and the output of the training sample, wherein the input of the training sample is the status of different devices in a certain scene, and the output of the training sample is the user's attention score for different devices.
  • the electronic device 100 uses the training data to iteratively train the attention degree calculation model based on the preset model to obtain a trained attention degree calculation model.
  • the training data may include states of different devices in multiple different scenarios and user attention scores for different devices.
  • the similarity between user C and user A using the mobile phone to turn on the do-not-disturb mode service and the mobile phone used by user A to turn on the do-not-disturb mode service is 0.1.
  • the similarity between the light-off service in the living room used by user C and the light-off service in the living room used by user A is 0.1.
  • the similarity between the service for turning on the air conditioner in the living room used by user C and the service for turning on the air conditioner in the living room used by user A is 0.
  • the similarity between the curtain closing service in the living room used by user C and the curtain closing service in the living room used by user A is 0.1.
  • the similarity between user D and user A using the mobile phone to turn on the do not disturb mode service and the mobile phone used by user A to turn on the do not disturb mode service is 1.
  • the similarity between the light-off service in the living room used by user D and the light-off service in the living room used by user A is 0.8.
  • the similarity between the service for turning on the air conditioner in the living room used by user D and the service for turning on the air conditioner in the living room used by user A is 0.3.
  • the similarity between the curtain closing service in the living room used by user D and the curtain closing service in the living room used by user A is 0.1.
  • the electronic device 100 recommends for user A the service sets used by similar users: the service set used by user D, the service set used by user B, and the service set used by user C.
  • the recommendation card 1600 generated by the electronic device 100 may not include the set of historical services used by user A, that is, the electronic device 100 does not display the card 1601 in the card 1600 .
  • the similarity between the call answering service used by user C and the call answering service used by user A is 0.
  • the similarity between the cockpit automatic recovery of the driving position service used by user C and the service of user A's automatic recovery of the cockpit position is 0.
  • the similarity between the navigation service used by user C and the navigation service used by user A is 0.3.
  • the similarity between the music playback service used by user C and the music playback service used by user A is 0.1.
  • a scenario is composed of scenario trigger conditions and service collections, and some scenarios may also include exit conditions.
  • the trigger condition is used to trigger the electronic device 100 to generate a recommendation card based on the service set corresponding to the current scene, and display the recommendation card on the desktop.
  • the exit condition is used to trigger the electronic device 100 to stop displaying the aforementioned recommendation card on the desktop.
  • Table 11 is only an example of the composition type and content of trigger conditions. In practical applications, it may include more composition types and content of trigger conditions than Table 11. The implementation of the present application Examples are not limited to this.
  • the exit condition of the morning scene may be composed of the above three conditions: the first condition is that one hour after the screen of the mobile phone is turned on for the first time after the screen of the mobile phone has been off for a long time.
  • the second condition is that the user is out of the house, and the location is getting farther and farther away from home. Only when the above two conditions are satisfied at the same time, the electronic device 100 will stop recommending the service set in the morning scene for the user.
  • the electronic device 100 Since the trigger conditions and exit conditions of the scene are predetermined, the user's usage habits are not uniform, which will cause the electronic device 100 to recommend a service set in a certain scene for the user according to the predetermined trigger conditions, or the electronic device 100 According to the predetermined exit conditions, stop recommending the service set in a certain scene for the user, and there is a deviation from the time when the user actually uses the service set in the scene. For example, the electronic device 100 recommends a service set in a certain scene for the user according to a predetermined trigger condition, which may cause the electronic device 100 to recommend a service set in a certain scene to the user too early, or the electronic device 100 recommends a certain service set for the user. It happens that the collection of services in a scene is too late.
  • the user will use the taxi service in the commuting scene on weekdays, and the triggering condition of the commuting scene is: after the user leaves home.
  • the user will use the taxi service to call a car before leaving home, and the user will leave home after the car is called. Due to the deviation between the preset trigger condition and the actual behavior of the user, the electronic device 100 cannot timely adjust the trigger condition of the scene and recommend corresponding services for the user.
  • the user in the scene of getting up in the morning, the user will use the speaker to play the weather service, and the triggering conditions of the scene of getting up in the morning are: 1. After 5:00 in the morning. 1. One hour after the screen is turned on for the first time after the phone has been off for a long time. 2. Within plus or minus one hour of the user's morning wake-up time. 3. After turning off the screen for a long time, the phone turns on for the first time. After the electronic device 100 detects that the time, location and device status meet the above trigger conditions in the morning, the electronic device 100 will recommend the speaker to play the weather service for the user.
  • the preset entry time of the scene is time t1
  • the preset exit time of the scene is time t2 .
  • the electronic device 100 recommends a service set in a certain scene to the user. Assuming that the user has not used the service set between time t1 and time t2, the electronic device 100 detects that the time reaches the preset exit time t2 of the scene. At time , the electronic device 100 will stop recommending the service set in the scene for the user. However, the user has not used the service set in the scene, and at a certain time after the electronic device 100 stops recommending the service set in the scene for the user, the user may want to use the service set in the scene again. In this way, the scene will be exited too early and the time when the user actually uses the service will be missed.
  • FIG. 19 exemplarily shows a functional block diagram of the electronic device 100 for adaptively adjusting scene conditions.
  • FIG. 21 exemplarily shows a schematic diagram of the electronic device 100 adjusting the trigger condition for entering a scene too early.
  • the maximum constraint condition in the preset trigger conditions can be understood as the last executed condition in the original trigger conditions.
  • the electronic device 100 records the behavior of the user within a period of time before and after using the service set in the scenario. For example, the electronic device 100 records the user's behavior for a period of time (for example, within 5 minutes) before the electronic device 100 recommends a set of services in the scene. The electronic device 100 records the behavior of the user within a period of time (for example, within 5 minutes) after the electronic device 100 stops the service set in the recommendation scenario.
  • the user's behavior can be understood as what the electronic device 100 detects the user does, or the electronic device 100 detects what the user does through other devices.
  • the electronic device 100 can add another constraint condition, the trigger condition of the morning scene is: 1. 7:15 in the morning, or the user's mobile phone turns on the screen for the first time within 7:15 in the morning. In this way, even if the time in the modified trigger condition is 7:00, but it is not yet 7:15, the user's mobile phone turns on and the user wakes up, and the electronic device 100 can recommend an alarm clock service for the user.
  • Table 12 shows user behaviors (or candidate new conditions) before the electronic device 100 recommends the service set of the scene in the morning scene. If there is one user behavior (or candidate addition condition) before the service set of the scene recommended by the electronic device 100, then this user behavior is used as an addition condition of the preset trigger condition. If there are multiple user behaviors prior to the service set of scenarios recommended by the electronic device 100, the electronic device 100 needs to determine an optimal user behavior from the multiple user behaviors as a new condition for the preset trigger condition.
  • the best candidate trigger condition is "1. After 5:00 in the morning. 2. 15 minutes after the phone is turned on for the first time after the screen has been off for a long time, or within 15 minutes after the phone is turned on for the first time after the screen has been off for a long time, The user finishes running and returns home", then the candidate added condition corresponding to the best candidate trigger condition is the added condition, that is, "completes running and returns home”.
  • the condition item modified relative to the preset trigger condition in the best candidate trigger condition is the maximum constraint condition, that is, "after the screen of the mobile phone is turned on for the first time after being off for a long time”.
  • modifying the trigger condition can be understood as the best candidate trigger condition obtained in S2102. Specifically, for how the electronic device 100 obtains the modified trigger condition, reference may be made to How to Obtain the Best Candidate Trigger Condition in S2102, which will not be repeated in this embodiment of the present application.
  • a scenario's exit condition can consist of one or more conditions. In the exit conditions of the scenario, find a condition that finally constrains the exit of the scenario, and this condition is the maximum constraint condition of the exit condition.
  • the electronic device 100 After the electronic device 100 determines the maximum constraint condition in the newly added condition and the preset exit condition, it modifies the maximum constraint condition in the preset exit condition based on the delay time ⁇ and the newly added condition, so that the modified maximum constraint condition can make the scenario
  • the exit time is later than the time actually used by the user, and the difference between the scene exit time and the time actually used by the user is within the preset value.
  • the aforementioned embodiments describe how the electronic device 100 generates scene services and how the electronic device 100 adjusts the entry conditions of the scene and the exit conditions of the scene.
  • the embodiment of the present application also provides another service verification method, that is, Say that after the electronic device 100 triggers a certain scene, the electronic device 100 will recommend a service set for the user, and generate a recommendation card based on the service set. Anomalies, such as incomplete display of service collection information in the recommendation card, that is, the display elements on the recommendation card are different from the default display elements. Or the recommendation card is displayed normally, but the user cannot directly use the service collection in the recommendation card after clicking the recommendation card.
  • Operations such as service recall, service filtering and sorting, and service collection generation on the electronic device 100 are all completed by the recommendation system on the electronic device 100 , and the recommendation system can be understood as a system application on the electronic device 100 .
  • the recommendation system determines that the recommended data of the home service does not include the location of the user's home, but the preset information in the verification rules preset by the home service includes the location of the user's home, then the recommendation system determines that The current home service cannot provide the user with a preset function, that is, to recommend route information from the current location to the user's home location for the user.
  • the recommendation system determines to enter a certain scene, and determines the service to be recommended to the user in this scene. Before the recommendation system recommends the service to be recommended to the user, the recommendation system needs to verify the service to be recommended.
  • the verification process includes the following steps:
  • the recommendation system obtains the identifier of the service to be recommended.
  • the service management module After receiving the identification of the service to be recommended sent by the recommendation system, the service management module sends a verification instruction to the service to be recommended.
  • the service management module sends the verification result to the recommendation system.
  • the recommendation system determines whether to recommend the service to be recommended to the user based on the verification result.
  • the service to be recommended is used to determine whether the preset function can be realized based on the current recommendation data and preset verification rules in response to the verification instruction after receiving the verification instruction, that is, to determine the verification result .
  • the verification result is ready.
  • the verification result is not ready.
  • Table 17 exemplarily shows the reputation scores corresponding to different services. It should be noted that the reputation score of the service to be recommended changes in real time. How to update the reputation score of the service to be recommended will be described in detail in subsequent embodiments, and will not be described in detail in this embodiment of the present application.
  • the service sorting module will determine the service to be recommended to the user from the plurality of services to be recommended based on the reputation score of the service to be recommended and the number of services finally pushed to the user.
  • the service sorting module After the service sorting module obtains the service finally recommended to the user, it sends the identification of the service finally recommended to the user to the service display module.
  • Fig. 30 exemplarily shows a schematic diagram of how to update the reputation score of the service to be recommended.
  • the card information collection module extracts the display content on the recommendation card, and the display content includes but not limited to text, icons, pictures and other information. Afterwards, the card information collection module sends the displayed content on the recommended card to the intent content matching module.
  • the intent content matching module determines the matching degree between the displayed content on the recommended card and the preset displayed content based on the extracted displayed content on the recommended card. If the matching degree is lower than the preset value, the intent content matching module determines that the displayed content on the recommendation card does not match the recommendation intent, and the intent content matching module sends the identification of the recommended service displayed on the recommendation card to the server. If the matching degree is higher than the preset value, the intent content matching module determines that the displayed content on the recommendation card matches the recommendation intent.
  • Service 1, Service 2, and Service 3 are ready after receiving the verification instruction, then the verification results sent by Service 1 and Service 3 to the service management module are ready. get ready. If the service 2 (identification of the first service function) is not ready, the verification result that the service 2 may send to the service management module is ready. Then the recommendation system may recommend the second service to the user.
  • the recommendation card of service two may be the recommendation card 2502 shown in FIG. 25A.
  • the recommendation card for service two can also be the recommendation card 2507 shown in FIG. 25B.
  • the service ranking module in the recommendation system will determine the final recommendation to the user from multiple services to be recommended based on the reputation score of the service to be recommended and the number of services that are finally pushed to the user. To be recommended for service.
  • the service sorting module determines that the service finally pushed to the user is Service 1 based on the reputation scores of Service 1, Service 2, and Service 3, and Service 1 can also be called For the second service function. If the number of services finally pushed to the user is 2, the service sorting module determines that the services finally pushed to the user are Service 1 and Service 3 based on the reputation scores of Service 1, Service 2, and Service 3, and Service 1 or Service 3 also Can be called the second service function.
  • the card information collection module extracts the displayed content on the recommended card and the intent content matching module determines the matching degree between the displayed content on the recommended card and the preset displayed content
  • the card information collection module will recommend the displayed content on the card.
  • the display content is sent to the server, and the intent content matching module sends the matching degree between the display content on the recommendation card and the preset display content to the server.
  • the server receives the display content on the recommendation card and the matching degree between the display content on the recommendation card and the preset display content
  • the server based on the display content on the recommendation card, the preset display content and the display content on the recommendation card and The preset matching degree of displayed content is used as sample data to train the intent content matching model.
  • the mobile phone sorts the above-mentioned multiple services, and determines which of the above-mentioned multiple services is most suitable to be recommended on the above-mentioned device.
  • the central control device can be determined from the multiple electronic devices that trigger the service recommendation process, and the central control device completes the tasks of service recall, service filtering, service reordering, and service loading matter.
  • Both the electronic device 100 and the electronic device 200 have a service recall module, a service filtering module, a service rough sorting module, a service reordering module, a service joint sorting module, a service filling module, and a device status and data management module.
  • the service filtering module sends the recalled services obtained after filtering to the service rough sorting module, and the service rough sorting module is used to sort the recalled services obtained after filtering according to simple rules, and obtain services whose user usage probability is greater than a preset value.
  • the service rough sorting module sends the identifiers of the recalled services whose user usage probability is greater than the preset value to the service reordering module, and the service reordering module will further sort the recalled services based on the user's usage habits. Screening to get the service of the eventual electronic device 100 recall.
  • the electronic device 200 also triggers the service recommendation process, then the service recall module, the service filtering module, the service rough sorting module and the service reordering module in the electronic device 100 and the service recall module in the electronic device 100,
  • the functions of the service filtering module, the service coarse sorting module and the service reordering module are similar, all of which are to obtain recalled services whose user usage probability is greater than a preset value.
  • the service reordering module in the electronic device 200 After the service reordering module in the electronic device 200 receives the service recalled by the final electronic device 200, the service reordering module in the electronic device 200 sends the service recalled by the final electronic device 200 to the device status and data management module in the electronic device 200 .
  • the service joint ranking module in the electronic device 100 After the service joint ranking module in the electronic device 100 obtains the service recalled by the electronic device 100 and the service recalled by the electronic device 200, the service joint ranking module in the electronic device 100 will base on the priority of the service, the user's current preference for the electronic device 100 and the electronic
  • the attention degree of the device 200, the service capacity on the electronic device 100 and the electronic device 200, and the matching degree of each service on the electronic device 100 and the electronic device 200 are used to sort the services recalled by the electronic device 100 and the electronic device 200, Get the optimal distribution of services on each device, and then distribute and transfer services, and finally display each service on the most suitable device to achieve cross-device recommendation services. How the service joint ordering module determines the optimal distribution of services on each device will be described in detail in subsequent embodiments, and will not be repeated here in this embodiment of the application.
  • the service joint ranking module can also obtain the correspondence degree score between each service and each electronic device based on other formulas, which is not limited in this embodiment of the present application.
  • the correspondence between Service 3 and mobile phones is A3, the correspondence between Service 3 and tablets is B3, the correspondence between Service 3 and large screens is C3, the correspondence between Service 3 and cars is D3, and the correspondence between Service 3 and watches is The corresponding score of the service three and the earphone is F3.
  • the correspondence between Service 4 and mobile phones is A4, the correspondence between Service 4 and tablets is B4, the correspondence between Service 4 and large screens is C4, the correspondence between Service 4 and cars is D4, and the correspondence between Service 4 and watches is The correspondence score of the service four and the earphone is F4. It should be noted that the correspondence scores between the above-mentioned services and each electronic device are constant.
  • ⁇ 1 , ⁇ 2 , ⁇ 3 , ⁇ 4 , ⁇ 5 and ⁇ 6 are all constants, and the specific explanations for ⁇ 1 , ⁇ 2 , ⁇ 3 , ⁇ 4 , ⁇ 5 and ⁇ 6 can be Refer to the relevant description in FIG. 34 above.
  • the car machine sends a request to the mobile phone, which requests the user to obtain the recalled service data and mobile phone status data on the mobile phone.
  • the mobile phone sends the mobile phone recall service to the car, namely express service and WeChat service.
  • the car machine will also combine the service capacity that can be accommodated on the car machine and the service capacity that can be accommodated on the mobile phone at the same time, and distribute each service to different devices.
  • Table 28 exemplarily shows the service capacity that can be displayed on each device at the same time. Among them, only one service can be displayed on the car at the same time, a maximum of three services can be displayed on the mobile phone, and only one service can be displayed on the headset at the same time.
  • the mobile phone is in the charging state; (2) The schedule starts in 3 hours; (3) The car is driving towards home, which is relatively close to home; (4) The car is using navigation software; (5) The user has never used WeChat and payment codes on the car.
  • the joint ranking result is that the usage probability of the express service is greater than that of the schedule service, the usage probability of the schedule service is greater than that of the WeChat service, and the usage probability of the WeChat service is greater than that of the navigation service.
  • the car-machine Based on Table 27 and Table 28, the car-machine finally determines that the correspondence between the express service and the car-machine has the highest score, and only one service can be displayed on the car-machine at the same time, then the car-machine will display the express service on the car-machine. Afterwards, the correspondence between the schedule service, WeChat service and navigation service and the mobile phone has the highest score, and a maximum of 3 services can be displayed on the mobile phone at the same time, then the vehicle will send the schedule service logo, WeChat service logo and navigation service logo to the mobile phone, The mobile phone displays the schedule service, WeChat service and navigation service on the mobile phone.
  • the historical use data of the first user may also be the user's historical operations. That is, the user behavior of the user within a period of time, such as the setting items of the first electronic device within a period of time, or the setting items of the first electronic device controlling other electronic devices. For example, yesterday the first user used the setting items of turning on the air conditioner and closing the curtains in the movie watching scene. Today, in the movie watching scene, the first electronic device recommends the recommendation card for the user to turn on the air conditioner and close the curtains.
  • the first electronic device executes multiple functional operations corresponding to the first service set.
  • the method further includes: , the first electronic device displays the first recommendation card corresponding to the first service set; when it is detected that the trigger condition of the first service set is met, the first electronic device executes multiple functional operations corresponding to the first service set, specifically including: When a trigger operation for the first recommended card is detected, the first electronic device executes multiple functional operations corresponding to the first service set.
  • the first recommendation card may be the card 801 shown in FIG. 9A .
  • the first recommendation card may also be the card 802 shown in FIG. 9C.
  • the first recommendation card may also be the card 803 shown in FIG. 9I.
  • the first recommendation card may also be the card 807 shown in FIG. 9J .
  • the first recommendation card may also be the card 810 shown in FIG. 9K .
  • the first recommendation card may also be the card 813 shown in FIG. 9L.
  • the first recommendation card may also be the card 1202 shown in FIG. 12 .
  • the first recommendation card may also be the card 1302 shown in FIG. 13 .
  • the first recommendation card may also be the card 1602 shown in FIG. 16 .
  • the first recommendation card may also be the card 1702 shown in FIG. 17 .
  • the first electronic device may simultaneously display recommendation cards corresponding to multiple service sets, and each service set may include multiple services, or may only include one service.
  • the first electronic device detects the user's trigger operation (for example, click) on the first recommended card, which enables the user to "use" multiple services with one click.
  • the user's trigger operation for example, click
  • the triggering condition can also be a voice command, the identity information of the first user, a gesture, etc., so that the first electronic device can be triggered to execute multiple services in the first service set without manual operation by the user, which is convenient and quick.
  • multiple service functions in the first service set correspond to multiple service functions of different applications, or multiple service functions in the first service set correspond to control of multiple different devices. In this way, different applications are controlled simultaneously or different devices are controlled simultaneously.
  • the multiple service function types in the first service set include any one or more of the following: application services, system services, and services for controlling working states of other electronic devices.
  • the first electronic device executes multiple functional operations corresponding to the first service set, specifically including: the first electronic device controls the first electronic device to work with a first working parameter, where the first working parameter includes any one or more of the following: The setting state of the system of the first electronic device, the setting state of the application installed on the first electronic device, and the opening/closing of the application installed on the first electronic device. And/or, the first electronic device controls the third electronic device to work with the second working parameter, wherein the second working parameter includes any one or more of the following: on/off of the third electronic device, setting state of the third electronic device .
  • the first electronic device generates a second service set based on the personal data of the first user.
  • the second service set includes multiple service functions, and the multiple functional operations corresponding to the second service set are the same as the multiple service functions corresponding to the first service set.
  • the functional operations are different; when it is detected that the trigger condition of the second service set is satisfied, the first electronic device executes multiple functional operations corresponding to the second service set. In this way, for the same user, in different scenarios, the first electronic device can automatically generate different service sets for the user, reflecting the intelligence of service recommendation.
  • the first electronic device generates a third service set based on the personal data of the second user, the third service set includes multiple service functions, and the personal data of the second user includes historical usage data of the second user;
  • the first electronic device executes multiple functional operations corresponding to the third service set. If the personal data of the first user is different from the personal data of the second user, the multiple functional operations corresponding to the third service set are different from the multiple functional operations corresponding to the first service set. In this way, in the same scene, for different users, the service sets generated by the electronic device are also different, which reflects the differences of users and the intelligence of service recommendation.
  • the first electronic device executes multiple functional operations corresponding to the first service set, specifically including: the first electronic device executes multiple functional operations corresponding to the first service set based on the execution relationship of multiple services in the first service set .
  • the execution relationship of multiple services in the first service set includes but not limited to sequential execution, simultaneous execution and so on.
  • the sequence of execution for example, in a navigation scenario, the first electronic device first acquires location information, and then acquires route information, and then the first electronic device broadcasts the route information through the earphone.
  • the method further includes: when the first condition is met, the first electronic device acquires multiple candidate service functions; the first electronic device Generating the first service set based on the personal data of the first user specifically includes: the first electronic device generating the first service set based on multiple candidate service functions based on the personal data of the first user. That is, before generating the first service set, the first electronic device needs to acquire multiple candidate service functions.
  • the first condition is the trigger condition of the scenario.
  • the method further includes: the first electronic device obtains a candidate service set, and the service functions in the candidate service set include the following Any one or more items: the service function sent by the server, the service function sent by other electronic devices, the service function preset in the first electronic device; the first electronic device obtains multiple candidate service functions, specifically including: the first electronic device The device acquires multiple candidate service functions from the candidate service set based on the personal data of the first user.
  • the multiple service sets are service sets whose similarity with the service function used by the first user in the first scenario is greater than a preset value; the first electronic device selects from the multiple service sets based on the personal data of the first user Determining the first service set specifically includes: the first electronic device determines the type of device that the first user cares about in the first scenario; Determine the first service set that is similar to the device type that the first user cares about from the service sets. That is to say, the first electronic device may recommend to the user a set of services used by similar users that are close to the type of device that the first user cares about.
  • the first electronic device obtains the first time when the user history uses the first service set and the second time when the first electronic device history displays the first recommended card; when the first electronic device determines that the first time is earlier than the second time time, or when the first time is later than the second time and the difference between the first time and the second time is greater than a preset value, the first electronic device adjusts the preset trigger condition of the first service set; After an electronic device adjusts the preset triggering condition of the first service set, the first electronic device displays the first recommendation card at a third time, wherein the third time is before the first time, and the third time and the first time The difference is less than the preset value.
  • the first electronic device adjusts the preset trigger condition of the first service set, which specifically includes: the first electronic device determines the maximum constraint condition of the preset trigger condition; the first electronic device determines the second new sub-set The condition, wherein the execution time of the second newly added subcondition is earlier than the execution time of the maximum constraint condition of the preset trigger condition; the first electronic device replaces the maximum constraint condition of the preset trigger condition with the second newly added subcondition.
  • the first electronic device can adjust the preset trigger condition that the scene is triggered too late.
  • the first electronic device can dynamically adjust the time when the first electronic device stops displaying the first service set based on the time when the user uses the first service set and the time when the user stops displaying the first service set, so as to prevent the first electronic device from stopping There is a large deviation between the time when the first service set is displayed and the time when the user uses the first service set.
  • the first electronic device adjusts the preset exit condition of the first service set, which specifically includes: the first electronic device determines the maximum constraint condition of the preset exit condition, and the preset exit condition includes one or more subconditions, The maximum constraint condition is the last executed subcondition among one or more subconditions; the first electronic device delays the execution time of the maximum constraint condition of the preset exit condition by a first duration;
  • the displaying of the first recommendation card by the first electronic device at the third time specifically includes: the first electronic device detects that the first new sub-condition is executed, and the first electronic device The first recommendation card is displayed at a third time.
  • the method further includes: before the third time, the first electronic device detects that the first newly added sub-condition is not executed but the maximum constraint condition of the preset exit condition is executed, The first electronic device displays the first recommendation card at a fourth time, and the fourth time is later than the third time.
  • a computer program product includes one or more computer instructions.
  • Computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, e.g. Coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) to another website site, computer, server or data center.
  • the processes can be completed by computer programs to instruct related hardware.
  • the programs can be stored in computer-readable storage media.
  • When the programs are executed may include the processes of the foregoing method embodiments.
  • the aforementioned storage medium includes: ROM or random access memory RAM, magnetic disk or optical disk, and other various media that can store program codes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • General Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Human Computer Interaction (AREA)
  • User Interface Of Digital Computer (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

La présente demande concerne un procédé de recommandation de service et un appareil associé. Le procédé consiste : à générer, par un premier dispositif électronique, un premier ensemble de services sur la base de données personnelles d'un premier utilisateur, le premier ensemble de services comprenant une pluralité de fonctions de service, et les données personnelles du premier utilisateur comprenant des données d'utilisation historiques du premier utilisateur (S3501) ; et lorsqu'il est détecté qu'une condition de déclenchement correspondant au premier ensemble de services est satisfaite, à exécuter, par le premier dispositif électronique, une pluralité d'opérations fonctionnelles correspondant au premier ensemble de services (S3502). Un premier dispositif électronique peut recommander automatiquement une pluralité de fonctions de service différentes à un utilisateur selon des données d'utilisation historiques de l'utilisateur, ce qui permet de réaliser la commodité et la précision de recommandation de service.
PCT/CN2022/142915 2021-12-30 2022-12-28 Procédé de recommandation de service et appareil associé WO2023125692A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111670318.7 2021-12-30
CN202111670318.7A CN116415061A (zh) 2021-12-30 2021-12-30 一种服务推荐方法及相关装置

Publications (1)

Publication Number Publication Date
WO2023125692A1 true WO2023125692A1 (fr) 2023-07-06

Family

ID=86997993

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/142915 WO2023125692A1 (fr) 2021-12-30 2022-12-28 Procédé de recommandation de service et appareil associé

Country Status (2)

Country Link
CN (1) CN116415061A (fr)
WO (1) WO2023125692A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117112116A (zh) * 2023-10-16 2023-11-24 成都市蓉通数智信息技术有限公司 基于数字政务的用户管理系统

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101414296A (zh) * 2007-10-15 2009-04-22 日电（中国）有限公司 自适应服务推荐设备及方法、自适应服务推荐系统及方法
WO2016048034A1 (fr) * 2014-09-23 2016-03-31 삼성전자 주식회사 Dispositif électronique et procédé de traitement d'informations du dispositif électronique
CN109829107A (zh) * 2019-01-23 2019-05-31 华为技术有限公司 一种基于用户运动状态的推荐方法及电子设备
CN111149124A (zh) * 2018-04-17 2020-05-12 华为技术有限公司 服务推荐方法及相关装置
CN111182145A (zh) * 2019-12-27 2020-05-19 华为技术有限公司 显示方法及相关产品
CN111466125A (zh) * 2018-08-03 2020-07-28 华为技术有限公司 服务推送方法及终端
CN111831903A (zh) * 2020-06-15 2020-10-27 北京嘀嘀无限科技发展有限公司 服务站点推荐方法、装置、设备及存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101414296A (zh) * 2007-10-15 2009-04-22 日电（中国）有限公司 自适应服务推荐设备及方法、自适应服务推荐系统及方法
WO2016048034A1 (fr) * 2014-09-23 2016-03-31 삼성전자 주식회사 Dispositif électronique et procédé de traitement d'informations du dispositif électronique
CN111149124A (zh) * 2018-04-17 2020-05-12 华为技术有限公司 服务推荐方法及相关装置
CN111466125A (zh) * 2018-08-03 2020-07-28 华为技术有限公司 服务推送方法及终端
CN109829107A (zh) * 2019-01-23 2019-05-31 华为技术有限公司 一种基于用户运动状态的推荐方法及电子设备
CN111182145A (zh) * 2019-12-27 2020-05-19 华为技术有限公司 显示方法及相关产品
CN111831903A (zh) * 2020-06-15 2020-10-27 北京嘀嘀无限科技发展有限公司 服务站点推荐方法、装置、设备及存储介质

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117112116A (zh) * 2023-10-16 2023-11-24 成都市蓉通数智信息技术有限公司 基于数字政务的用户管理系统
CN117112116B (zh) * 2023-10-16 2024-02-02 成都市蓉通数智信息技术有限公司 基于数字政务的用户管理系统

Also Published As

Publication number Publication date
CN116415061A (zh) 2023-07-11

Similar Documents

Publication Publication Date Title
WO2021018154A1 (fr) Procédé et appareil de représentation d'informations
US8886576B1 (en) Automatic label suggestions for albums based on machine learning
US8429103B1 (en) Native machine learning service for user adaptation on a mobile platform
US8510238B1 (en) Method to predict session duration on mobile devices using native machine learning
CN110727638B (zh) 电子系统中的数据系统和数据调度方法以及机器可读介质
KR102446811B1 (ko) 복수의 디바이스들로부터 수집된 데이터 통합 및 제공 방법 및 이를 구현한 전자 장치
US20190057298A1 (en) Mapping actions and objects to tasks
US20130346347A1 (en) Method to Predict a Communicative Action that is Most Likely to be Executed Given a Context
US8560515B2 (en) Automatic generation of markers based on social interaction
CN106537946A (zh) 对用于移动设备唤醒的信标消息进行评分
EP3627806A1 (fr) Procédé de génération de portrait d'utilisateur, et terminal
US20150095158A1 (en) Security for mobile system
US10464570B2 (en) Electronic device and method using machine learning for identifying characteristics of users located within a specific space
CN113722581B (zh) 一种信息推送方法和电子设备
WO2019140702A1 (fr) Procédé et dispositif permettant de générer une image de profil d'utilisateur
CN116126509A (zh) 基于多设备提供服务的方法、相关装置及系统
WO2023125692A1 (fr) Procédé de recommandation de service et appareil associé
CN106228511A (zh) 电子装置以及用于在电子装置中产生图像文件的方法
CN113495487A (zh) 一种调节目标设备运行参数的终端和方法
CN113312543A (zh) 基于联合学习的个性化模型训练方法、电子设备和介质
WO2024032707A1 (fr) Procédé de recommandation de service et dispositif électronique
WO2023179490A1 (fr) Procédé de recommandation d'application et dispositif électronique
WO2022057764A1 (fr) Procédé d'affichage de publicité et dispositif électronique
CN115809362A (zh) 内容推荐方法及电子设备
CN113762585B (zh) 数据的处理方法、账号类型的识别方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22914943

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE