Information optimization recommendation method and device, electronic equipment and computer readable storage medium

Abstract

The invention provides an information optimization recommendation method and device, and relates to the technical field of computers, in particular to the technical fields of large models, deep learning, natural language processing and the like. The specific implementation scheme is as follows: collecting operation behavior data of operating a target application by a user; determining initial function information of the target application based on the operation behavior data and historical control data of the user; determining an initial operation sequence based on the initial function information, and sending initial prompt information of the initial operation sequence; receiving feedback data of the user; based on the feedback data and the initial operation sequence, the recommended operation sequence is obtained, and the user experience is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computers, specifically to technical fields such as large models, deep learning, and natural language processing. In particular, it relates to a method and device for optimizing information recommendation, an electronic device, a computer-readable storage medium, and a computer program product. Background Art

[0002] With the continuous enrichment of application functions on mobile terminals, users often need to find the functions they commonly use or have potential needs among numerous functions. However, the existing function organization method has a deep hierarchy, and users need to click multiple times to reach the recommended functions, resulting in a poor user experience. At the same time, traditional big data recommendation methods mainly rely on server-side computing, making it difficult to achieve personalized recommendations for end users and having privacy and storage cost issues. Summary of the Invention

[0003] The present disclosure provides a method and device for optimizing information recommendation, an electronic device, a computer-readable storage medium, and a computer program product.

[0004] According to a first aspect, there is provided a method for optimizing information recommendation, the method including: collecting operation behavior data of a user operating a target application; determining initial function information of the target application based on the operation behavior data and the user's historical manipulation data; determining an initial operation sequence based on the initial function information, and sending an initial prompt message of the initial operation sequence; receiving feedback data of the user; and obtaining a recommended operation sequence based on the feedback data and the initial operation sequence.

[0005] According to a second aspect, there is provided an apparatus for optimizing information recommendation, the apparatus including: a collection unit configured to collect operation behavior data of a user operating a target application; a determination unit configured to determine initial function information of the target application based on the operation behavior data and the user's historical manipulation data; a sending unit configured to determine an initial operation sequence based on the initial function information, and send an initial prompt message of the initial operation sequence; a receiving unit configured to receive feedback data of the user; and a obtaining unit configured to obtain a recommended operation sequence based on the feedback data and the initial operation sequence.

[0006] According to a third aspect, there is provided an electronic device, the electronic device including: at least one processor; and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method described in any implementation manner of the first aspect.

[0007] According to a fourth aspect, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the method described in any implementation manner of the first aspect.

[0008] According to a fifth aspect, there is provided a computer program product including a computer program which, when executed by a processor, implements the method described in any implementation manner of the first aspect.

[0009] For the information optimization recommendation method and apparatus provided in the embodiments of the present disclosure, first, operation behavior data of a user operating a target application is collected; second, based on the operation behavior data and the user's historical operation data, initial function information of the target application is determined; third, based on the initial function information, an initial operation sequence is determined, and an initial prompt message of the initial operation sequence is sent; fourth, feedback data of the user is received; and finally, based on the feedback data and the initial operation sequence, a recommended operation sequence is obtained. Thus, on the terminal side, based on the operation behavior data of the user, the initial operation sequence corresponding to the initial function information is determined, and the user can be effectively reminded of the functions that can be operated through the initial operation sequence, improving the user's function usage efficiency; based on the user's feedback data and the initial operation sequence, the recommended operation sequence is determined, improving the accuracy and reliability of the recommended operation sequence recommended to the user.

[0010] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0012] Figure 1 is a flowchart of an embodiment of the information optimization recommendation method according to the present disclosure;

[0013] Figure 2 is a schematic structural diagram of various data interactions and storages in the information optimization recommendation process of the present disclosure;

[0014] Figure 3 is a flowchart of another embodiment of the information optimization recommendation method of the present disclosure;

[0015] Figure 4 is a flowchart of still another embodiment of the information optimization recommendation method of the present disclosure;

[0016] Figure 5 is a schematic structural diagram of an embodiment of the information optimization recommendation apparatus according to the present disclosure;

[0017] Figure 6It is a block diagram of an electronic device used to implement the information optimization recommendation method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0018] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0019] In view of the defects in the traditional technology, the present invention proposes an information optimization recommendation method, which can effectively recommend a recommended operation sequence for users, so that users can effectively use the functions of the application. Figure 1 A process 100 of an embodiment of the information optimization recommendation method according to the present disclosure is shown. The information optimization recommendation method is applied to the terminal side. The information optimization recommendation method includes the following steps:

[0020] Step 101: Collect operation behavior data of a user operating a target application.

[0021] In this embodiment, the operation behavior data is the data of the user's operation of the target application, and the operation behavior data includes but is not limited to data such as clicks, slides, browses, browse times, stay times, interaction sequence, etc. The operation behavior data may also include: operations of the interaction sequence of different controls on the target application in a set time period. By analyzing the user's behavior on the operation behavior data and extracting the user's preference pattern, it is possible to avoid the target recommendation sequence being inaccurate due to one-sided data.

[0022] In this embodiment, data tracking points can be preset on the target application installed on the terminal side through the full embedding technology, without or with only a small amount of code intervention, that is, the user's operation behavior data on all control operations on the target application can be automatically collected. The implementation principle of full embedding mainly relies on technologies such as annotation processors. The annotation processor scans and processes annotations during compilation and generates corresponding codes to achieve automatic embedding. Specifically, the effect of automatic embedding is achieved by intercepting or inserting the original processing logic of the control of the target application.

[0023] In this embodiment, the operation behavior data automatically recorded includes but is not limited to: function usage frequency data (the number of times a function is used and the interval time), browsing behavior data (the length of time a page stays and the number of scrolls), interactive behavior data (interaction methods such as clicking buttons, sliding, and inputting), and operation paths (the steps and sequence of user access to functions). Since the information optimization recommendation method runs on the terminal side, the operation behavior data is stored in the terminal database on the terminal side, such as Figure 2As shown, it is not uploaded to the server to ensure privacy security.

[0024] In this embodiment, when collecting operation behavior data, it is carried out in a non-invasive manner, which does not affect the user experience, and the collection granularity can be dynamically adjusted to ensure the accuracy and comprehensiveness of the operation behavior data.

[0025] Optionally, after collecting the operation behavior data, the operation behavior data can be stored in the form of data encryption to protect user privacy and prevent unauthorized access and abuse.

[0026] In this embodiment, the operation behavior data is stored locally on the terminal device and is managed using a lightweight database (such as SQLite, PostgreSQL, etc.) or an efficient storage format (such as protobuf, JSON, etc.). When storing the operation behavior data, data compression and data cleaning strategies are adopted for the stored operation behavior data. The data compression and data cleaning strategies include: setting a data expiration time and limiting the maximum storage capacity. Through the data compression and data cleaning strategies, storage occupancy and performance overhead can be reduced.

[0027] In the technical solution of the present disclosure, the processing of the collection, storage, use, processing, transmission, provision, and disclosure of the involved operation behavior data is carried out after authorization and complies with relevant laws and regulations. The information related to the user in the operation behavior data is obtained after the user's permission, and the information related to the user in the operation behavior data is processed under confidentiality conditions.

[0028] Step 102: Determine the initial function information of the target application based on the operation behavior data and the user's historical manipulation data.

[0029] In this embodiment, the user's historical manipulation data is the data obtained after the user operates on the functions of the target application in the historical time period. The historical manipulation data includes: historical behavior data and the functions corresponding to the historical behavior data. The historical behavior data is the operation behavior data of the user in the historical time period. The historical behavior data includes at least one operation sequence, and each operation sequence corresponds to a function. The function corresponding to the historical behavior data is the function corresponding to all operation sequences. The type of the function corresponding to the historical behavior data can be determined by the type of functions implemented by the target application, and the number of the functions corresponding to the historical behavior data is determined by the user's preference. By analyzing the currently collected operation behavior data and the historically collected historical manipulation data, the user's function operation requirements and function operation preferences can be effectively analyzed, so as to effectively recommend functions for the user.

[0030] In this embodiment, the processing of the collection, storage, use, processing, transmission, provision, and disclosure of the historical manipulation data, etc., is performed after authorization and complies with relevant laws and regulations. The information related to the user in the historical manipulation data is the information obtained after the user's permission, and the information related to the user in the historical manipulation data can be processed under confidentiality conditions.

[0031] In this embodiment, the initial function information of the target application is the function information that can be initially recommended to the user. The initial function information includes: function identifier, function name, function description, etc. The above step 102 includes: determining a recommended operation from the operation behavior data and the historical manipulation data; based on the operation behavior data and the historical manipulation data, detecting whether the number of operations or the operation duration of the user for the recommended operation reaches the corresponding operation threshold; in response to detecting that the number of operations or the operation duration for the recommended operation reaches the operation threshold, taking the function information corresponding to the recommended operation as the initial function information. For example, if the number of views, the stay time, and the number of clicks of a certain function reach a certain threshold, taking the function information of this function as the initial function information.

[0032] Step 103, based on the initial function information, determine an initial operation sequence and send an initial prompt message for the initial operation sequence.

[0033] In this embodiment, the initial operation sequence is the operation sequence that the execution entity on which the information optimization recommendation method runs is to recommend to the user. The initial operation sequence is a sequence related to the initial function information. The initial operation sequence corresponds to the initial function, and the initial function can be a function of a target application or multiple functions in the target application.

[0034] In this embodiment, the initial prompt message is the information for prompting the user that they can select the operation sequence in the initial operation sequence. The initial prompt message includes: the initial operation sequence, the initial function, and the priority levels of each operation sequence in the initial operation sequence. Among them, the initial function is the function that the execution entity on which the information optimization recommendation method runs recommends to the user, the initial function information is the information of the initial function, the initial operation sequence includes at least one operation sequence, and each operation sequence can have a corresponding priority level. The priority level is the level of the priority recommendation information.

[0035] In this embodiment, each function of the target application corresponds to a corresponding operation sequence. The operation sequence is a sequence including at least one operation. The operations in the at least one operation can be actions of manipulating units (such as components or controls, etc.) in the target application (such as clicking a button), or actions of input units of input devices (such as keyboards, mice) (such as sliding a mouse). The initial function information is the information of the initial function. By querying the operation sequence corresponding to the initial function, the initial operation sequence is obtained. It should be noted that the initial function information is the recommended information of the current round of operation behavior data. By matching the operation behavior data with the recommended operation sequence of the previous round, the recommended operation sequence of the previous round can be included in the initial function information. Thus, the initial operation sequence can be the recommended operation sequence of the previous round.

[0036] Optionally, the above steps include: determining the initial function based on the initial function information; determining the operation sequence corresponding to the initial function, and in response to the operations in the operation sequence matching the operation data in the operation behavior data, starting from the operation data and connecting all the subsequent operations in the operation sequence to obtain the initial operation sequence, and sending the initial prompt information of the initial operation sequence.

[0037] It should be noted that the initial operation sequence can include one operation sequence or multiple operation sequences, and each operation sequence corresponds to one function.

[0038] Step 104, receiving the feedback data of the user.

[0039] In this embodiment, the execution entity on which the information optimization recommendation method runs can display the initial prompt information of the initial operation sequence in the preset display area of the target application, and receive the feedback data of the user through this display area.

[0040] In this embodiment, the feedback data is the feedback result of the user on the initial operation sequence. The feedback data includes: explicit feedback data and implicit feedback data. Among them, the display feedback data is the direct feedback result of the user on the initial operation sequence. For example, the explicit feedback data includes: like or dislike, adjusting the priority of the operation sequence or operation, etc. The implicit feedback data is the indirect feedback result of the user on the initial operation sequence. For example, clicking the initial function in the initial prompt information, or the frequency of operating the controls of the target application changes, etc.

[0041] In this embodiment, the feedback data of the user can be received through the terminal device operated by the user, or the feedback data can be determined by monitoring the operations of the user on the controls or components of the target application.

[0042] Step 105, obtaining the recommended operation sequence based on the feedback data and the initial operation sequence.

[0043] In this embodiment, the recommended operation sequence is an operation sequence that meets the user's functional requirements. This recommended operation sequence can directly be the initial operation sequence when the feedback data belongs to some situations, and when the feedback data belongs to other situations, the recommended operation sequence is an operation sequence different from the initial operation sequence. After obtaining the user's recommended operation sequence through the initial operation sequence, the recommended operation sequence can be associated with the initial function information to obtain historical manipulation data.

[0044] The above step 105 includes: in response to detecting that the operation frequency of the user operating the predetermined function in the feedback data is greater than the frequency threshold, determining the operation sequence of the predetermined function, and using the operation sequence of the predetermined function as the target sequence, where the predetermined function is the function that the user frequently operates as characterized by the feedback data.

[0045] The information optimization recommendation method provided by the embodiment of the present disclosure includes: First, collect the operation behavior data of the user operating the target application; Second, based on the operation behavior data and the user's historical manipulation data, determine the initial function information of the target application; Third, based on the initial function information, determine the initial operation sequence and send the initial prompt information of the initial operation sequence; Fourth, receive the feedback data of the user; Finally, based on the feedback data and the initial operation sequence, obtain the recommended operation sequence. Thus, on the terminal side, based on the user's operation behavior data, the initial operation sequence corresponding to the initial function information is determined, which can effectively remind the user of the functions that can be operated through the initial operation sequence, improving the user's functional usage efficiency; based on the user's feedback data and the initial operation sequence, the recommended operation sequence is determined, improving the accuracy and reliability of the recommended operation sequence recommended to the user.

[0046] Figure 3 Flow 300 showing another embodiment of the information optimization recommendation method according to the present disclosure is shown. The above information optimization recommendation method includes the following steps:

[0047] Step 301, collect the operation behavior data of the user operating the target application.

[0048] Step 302, based on the operation behavior data and the user's historical manipulation data, detect whether the function recommendation condition is met; if it is detected that the function recommendation condition is met, then execute step 303; if it is detected that the function recommendation condition is not met, then execute step 301.

[0049] In this embodiment, the function recommendation condition is the condition for detecting whether to perform function recommendation. When the execution entity on which the information optimization recommendation method runs detects that the function recommendation condition is met through operation behavior data and historical manipulation data, it is determined that the initial function information can be pushed to the user. To facilitate the user's direct understanding of the initial function information, the execution entity will convert the initial function information into an initial prompt message including an initial operation sequence and send it to the user. If the execution entity detects that the function recommendation condition is not met, the initial prompt message will not be sent to the user.

[0050] In this embodiment, the function recommendation conditions include: the time taken for the operation sequence corresponding to the operation behavior data is much less than the time taken for the same operation sequence in the historical manipulation data, and it is determined that the function corresponding to the operation sequence is a high-frequency function. For example, when the user browses photos, the user's habit may be to frequently click on the control of the preview function in the photo album. At this time, the photo album should be recommended with high priority instead of other positions where the photos are located.

[0051] Optionally, the function recommendation conditions include that at least one of the device type, application usage scenario, and past behavior pattern in the operation behavior data and the historical manipulation data is the same as the user's portrait data.

[0052] Step 303: Based on the operation behavior data and the historical manipulation data, determine the initial function information of the target application.

[0053] As Figure 2 shown, after detecting that the function recommendation condition is met, first obtain the functions of the target application and determine the initial function information.

[0054] Step 304: Based on the initial function information, determine the initial operation sequence and send the initial prompt message of the initial operation sequence.

[0055] As Figure 2 shown, after determining the initial function information, determine the initial operation sequence of the initial function and send the initial prompt message of the initial operation sequence.

[0056] Step 305: Receive the feedback data of the user.

[0057] Step 306: Based on the feedback data and the initial operation sequence, obtain the recommended operation sequence.

[0058] As Figure 2 shown, after sending the initial prompt message, based on the feedback data of the user, obtain the recommended operation sequence.

[0059] It should be understood that the operations and features in the above-mentioned step 301 and steps 303 - 306 respectively correspond to the operations and features in steps 101 - 105. Therefore, the descriptions of the operations and features in steps 101 - 105 above also apply to step 301 and steps 303 - 306, and will not be elaborated here.

[0060] For the information optimization recommendation method provided in this embodiment, after collecting operation behavior data, it detects whether the function recommendation condition is met based on the operation behavior data and the user's historical manipulation data; in response to detecting that the function recommendation condition is met, it determines the initial function information of the target application based on the operation behavior data and the historical manipulation data, providing a reliable implementation method for obtaining the initial function information.

[0061] In some embodiments of the present disclosure, the detecting whether the function recommendation condition is met based on the operation behavior data and the user's historical manipulation data includes: determining the total number of operations of each function of the target application based on the operation behavior data and the user's historical manipulation data; detecting whether the total number of operations is greater than or equal to a first set threshold; in response to detecting that the total number of operations is greater than or equal to the first set threshold, determining that the function recommendation condition is met.

[0062] In this alternative implementation, the target application includes multiple functions, such as a call function, a video function, etc. When the user uses the functions of the target application, an operation sequence for the function needs to be operated. For example, when the user needs to make a call through the target application, they need to first click on the component to return to the home page, and then click on the call control button to implement the call function.

[0063] In this alternative implementation, the total number of operations is the total number of operation sequences corresponding to the function, such as the total number of operation sequences for the call function. The first set threshold can be set according to development requirements. For example, the first set threshold is 3 times.

[0064] In this alternative implementation, the operation behavior data and the historical manipulation data may have operation sequences of the user for the functions. The number of operation sequences of each function in the operation behavior data and the historical manipulation data is accumulated to obtain the total number of operations of each function, and it is detected whether the total number of operations of each function is greater than or equal to the first set threshold to determine that the function recommendation condition is met.

[0065] The method for detecting whether the function recommendation condition is met provided by this alternative implementation determines the total number of operations of each function of the target application based on the operation behavior data and the user's historical manipulation data; detects whether the total number of operations is greater than or equal to the first set threshold; in response to detecting that the total number of operations is greater than or equal to the first set threshold, it is determined that the function recommendation condition is met. By detecting the total number of operations of each function to determine whether the function recommendation condition is met, the comprehensiveness of the detection of the current operation state is improved.

[0066] Optionally, the above detection of whether the function recommendation condition is met based on the operation behavior data and the user's historical manipulation data includes: determining the operation frequency of each function of the target application based on the operation behavior data and the user's historical manipulation data; detecting whether the operation frequency is greater than or equal to the frequency set value; in response to detecting that the operation frequency is greater than or equal to the frequency set value, it is determined that the function recommendation condition is met.

[0067] In some alternative implementations of the present disclosure, the above detection of whether the function recommendation condition is met based on the operation behavior data and the user's historical manipulation data further includes: in response to detecting that the total number of operations is less than the first set threshold, determining the segmented operation number of the function in the time window based on the operation behavior data and the historical manipulation data; detecting whether the segmented operation number is greater than or equal to the second set threshold, where the second set threshold is less than the first set threshold; in response to detecting that the segmented operation number is greater than or equal to the second set threshold, it is determined that the function recommendation condition is met.

[0068] In this alternative implementation, the time corresponding to the operation behavior data and the historical manipulation data is divided into multiple time windows, and the time periods of each time window are the same. By judging the segmented operation number in the time window, it can be determined whether the function is a high-frequency function. The specific time value of the time window can be determined based on the development requirements. For example, the time window is within 1 s; the segmented operation number refers to the number of times the function is operated in the time window, and the operation frequency of the function can be determined through the segmented operation number. The second set threshold can be set according to the development requirements, such as the second set threshold is 2 times.

[0069] The method for detecting whether the function recommendation condition is met provided by this alternative implementation, in response to detecting that the total number of operations is less than the first set threshold, determines the segmented operation number of the function in the time window based on the operation behavior data and the historical manipulation data; detects whether the segmented operation number is greater than or equal to the second set threshold, where the second set threshold is less than the first set threshold; in response to detecting that the segmented operation number is greater than or equal to the second set threshold, it is determined that the function recommendation condition is met. After the total number of operations of the function does not meet the requirements, the operation frequency of the function is judged by detecting the segmented operation number in the time window, which improves the comprehensiveness of the detection of the current operation state.

[0070] In some alternative implementations of the present disclosure, in response to detecting that the function recommendation condition is satisfied, based on the operation behavior data and the historical manipulation data, determining the initial function information of the target application includes: in response to detecting that the function recommendation condition is satisfied, inputting the operation behavior data and the historical manipulation data into a function recommendation model, and obtaining the initial function information output by the function recommendation model, where the function recommendation model is installed on the terminal side.

[0071] In this alternative implementation, the function recommendation model is a pre-trained model. After inputting the user's operation behavior data combined with the historical manipulation data into the function recommendation model, the function recommendation model outputs the initial function information. Among them, the historical manipulation data includes the historical initial function information, and the function recommendation model can output the optimal initial function information in combination with the historical initial function information.

[0072] In this alternative implementation, the function recommendation model can perform information optimization recommendation based on content. Specifically, a deep learning model (such as a convolutional neural network) is used to analyze the function features (such as function descriptions, parameters, usage scenarios, etc.) and user preference features (such as features extracted from the user's historical operations, browsing records, etc.) in the operation behavior data and the historical manipulation data. By learning the relationships between these features, functions related to the user's preferences are recommended to the user.

[0073] Optionally, the function recommendation model can perform information optimization recommendation through collaborative filtering. Specifically, the function recommendation model uses a multi-layer neural network in deep learning to learn the potential relationships between users and functions in the operation behavior data and the historical manipulation data. Users and functions can be represented as vectors, and the neural network is used to predict the user's interest level in functions that have not been used.

[0074] Optionally, the function recommendation model can also perform information optimization recommendation in the way of sequence modeling. Specifically, a recurrent neural network (RNN) or a long short-term memory network (LSTM) is used to process the user's operation sequence data. The function recommendation model can capture the time sequence and patterns of the user's behavior, so as to predict the next function that may be of interest.

[0075] Optionally, the function recommendation model can also perform information optimization recommendation in the way of multi-modal data fusion. Specifically, the function recommendation model combines various types of data, such as images, texts, audios, etc., to comprehensively understand functions and users. For example, for a function with image and text descriptions, a deep learning model can be used to process these two modalities of data simultaneously to provide more accurate recommendations.

[0076] Optionally, the function recommendation model can also adopt reinforcement learning for information optimization recommendation, regarding the recommendation system as a process of interaction between an agent and the environment. Using the user's feedback (such as clicks, usage duration, evaluations, etc.) as the reward signal, the model is trained to optimize the recommendation strategy, enabling it to continuously improve the recommendation effect over time.

[0077] In this optional implementation, when the function recommendation model adopts models with different structures, it can be trained using the training methods of the models with such structures.

[0078] Optionally, the function recommendation model can be trained using a pre-trained model and the principle of transfer learning. Specifically, a pre-trained deep learning model (such as a language model) on a large amount of data is utilized, and it is fine-tuned on specific function recommendation tasks to utilize the general knowledge and feature representations that have been learned.

[0079] The method for obtaining initial function information provided by this optional implementation, when it detects that the function recommendation condition is met, inputs the operation behavior data and historical manipulation data into the function recommendation model to obtain the initial function information output by the function recommendation model. The function recommendation model is installed on the terminal side. Through the function recommendation model located on the terminal side, the operation behavior data and historical manipulation data are analyzed to obtain the initial function information, improving the reliability of obtaining the initial function information.

[0080] Figure 4 Flow 400 of another embodiment of the information optimization recommendation method according to the present disclosure is shown. The above information optimization recommendation method includes the following steps:

[0081] Step 401, collect the operation behavior data of the user operating the target application.

[0082] As Figure 2 shown, after storing the operation behavior data, obtain the server configuration information. The server configuration information can include various information, such as the deployment information of the function recommendation model and the function recommendation condition. Through the deployment information in the server configuration information, the function recommendation model can be deployed on the terminal side, and through the function recommendation condition, it can be detected whether to perform the initial function information optimization recommendation.

[0083] Step 402, based on the operation behavior data and the user's historical manipulation data, detect whether the function recommendation condition is met; if it is detected that the function recommendation condition is met, then execute step 403; if it is detected that the function recommendation condition is not met, then execute step 401.

[0084] Step 403, input the operation behavior data and historical manipulation data into the function recommendation model to obtain the initial function information output by the function recommendation model.

[0085] In this embodiment, the function recommendation model is installed on the terminal side. For the specific implementation principle and training process of the function recommendation model, reference can be made to the description of the above optional implementation manners.

[0086] Optionally, the function recommendation model can also extract user habit features and function features from operation behavior data and historical manipulation data, combine the time dimension (frequency change in the most recent 7 days / 30 days), identify whether the user habits corresponding to the user habit features have changed, calculate the usage popularity scores of the functions corresponding to each function feature, sort them from high to low, and form initial personalized function information. For similar functions recommended in the initial function information, if a certain function is used frequently, the system can recommend functions with higher relevance.

[0087] Step 404: Based on the initial function information, determine an initial operation sequence and send an initial prompt message for the initial operation sequence.

[0088] Step 405: Receive the feedback data of the user.

[0089] Step 406: Based on the feedback data, adjust the weights and parameters of the function recommendation model and send the weights and parameters to the server.

[0090] In this embodiment, through the feedback data of the user, the functions frequently used by the user and the functions rarely used by the user can be determined. For the functions frequently used, the parameters and weights of the function recommendation model are improved; for the functions rarely used but related to the frequently used functions, the weights are also appropriately increased.

[0091] In this embodiment, sending the weights and parameters to the server enables the server to retrain the function recommendation model and use the trained function recommendation model again to generate the initial function information.

[0092] Step 407: Based on the feedback data and the initial operation sequence, obtain a recommended operation sequence.

[0093] It should be understood that the operations and features in the above steps 401, 402, 404 - 405, and 407 respectively correspond to the operations and features in steps 301, 302, 304 - 306. Therefore, the descriptions of the operations and features in steps 301, 302, 304 - 306 above also apply to steps 401, 402, 404 - 405, and 407, and will not be repeated here.

[0094] The information optimization recommendation method provided in this embodiment includes the following stages. The first stage: data accumulation. When the user starts using the target application, the system collects their operation behaviors in the background. The second stage: recommendation trigger. After a period of time, the system provides personalized recommendations in areas such as the home page, quick access, and search bar. The third stage: user feedback. The user can accept the recommendation, manually adjust it, or ignore it. The fourth stage: recommendation optimization. The function recommendation model dynamically optimizes the recommendation weights based on user feedback to improve the recommendation quality. The fifth stage: continuous iteration. The system regularly updates the recommendation results to ensure they meet the user's current needs.

[0095] The information optimization recommendation method provided in this embodiment, in response to detecting that the function recommendation condition is met, inputs the operation behavior data and historical manipulation data into the function recommendation model to obtain the initial function information output by the function recommendation model. After obtaining the user's feedback data, based on the feedback data, adjusts the weights and parameters of the function recommendation model, and sends the weights and parameters to the server. By adjusting the weights and parameters, the function recommendation model is effectively updated. When generating the initial function information through the function recommendation model subsequently, the initial function information can be made more accurate, improving the reliability of the obtained target recommendation information.

[0096] In some embodiments of the present disclosure, the above information optimization recommendation method further includes: determining the user's operation habit data based on the historical manipulation data and operation behavior data; determining the user's emotional state based on the operation habit data; and inputting the emotional state into the function recommendation model.

[0097] In this embodiment, the user operation habit data refers to the data of the fixed behavior patterns and operation habits formed by the user in operating the target application. These data are collected through data burying points, recording various behaviors of the user on the platform, such as clicking, browsing, swiping, long pressing, etc. For some operations

[0098] In this embodiment, by analyzing some operation characteristics (such as operation speed, click strength, etc.) in the operation habit data, the user's emotional state (such as happy, angry, anxious, etc.) can be directly determined.

[0099] In this embodiment, the function recommendation model can be a large model. Inputting the user's emotional state into the function recommendation model can enable the function recommendation model to determine the user's current emotional state, thereby giving better initial function information.

[0100] The information optimization recommendation method provided in this embodiment, through the historical manipulation data and operation behavior data, determines the user's operation habit data; determines the user's emotional state based on the operation habit data; and inputs the emotional state into the function recommendation model, which can enable the function recommendation model to recommend the initial function information based on the user's current emotional state, improving the comprehensiveness of the obtained initial function information.

[0101] Optionally, the above information optimization recommendation method further includes: determining user portrait data of a user based on operation behavior data and the user's historical manipulation data; inputting the user portrait data into a function recommendation model, which can enable the function recommendation model to recommend initial function information based on the user's current user portrait data, improving the comprehensiveness of the obtained initial function information. The function recommendation model can provide differentiated recommendation strategies according to different user portraits. For example, for new users, the function recommendation model recommends core basic functions to reduce the learning cost. For old users, the function recommendation model predicts the functions that may be needed based on historical behaviors to optimize the usage path. For highly active users: more emphasis is placed on recommending quick operations to improve operation efficiency.

[0102] In some optional implementation manners of the present disclosure, the detecting whether the function recommendation condition is satisfied based on the operation behavior data and the user's historical manipulation data includes: determining the user portrait data of the user based on the operation behavior data and the user's historical manipulation data; detecting whether the user portrait data matches the library portrait data of the corresponding function in a preset portrait database; and in response to detecting that the user portrait data matches the library portrait data, determining that the function recommendation condition is satisfied.

[0103] In this optional implementation manner, the user portrait data refers to characterizing the feature attributes of a user or a product by collecting and analyzing various dimension data of the user, including social attributes, consumption habits, preference characteristics, etc., and performing analysis, statistics, and mining of potential value information, so as to abstract the overall picture of the user's information.

[0104] In this optional implementation manner, the portrait database is a database corresponding to portraits and functions set for multiple users. The portrait database includes the library portrait data of each user and the historical operation functions corresponding to each library portrait data. Calculate the similarity value between the user portrait data and the library portrait data in the portrait database. When it is detected that the similarity value is greater than the similarity threshold, it is determined that the user portrait data matches the library portrait data. At this time, the historical operation function corresponding to the library portrait is the function corresponding to the library portrait.

[0105] The method for detecting whether the function recommendation condition is satisfied provided by this optional implementation manner determines the user portrait data of the user based on the operation behavior data and the user's historical manipulation data; detects whether the user portrait data matches the library portrait data of the corresponding function in a preset portrait database; and in response to detecting that the user portrait data matches the library portrait data, determines that the function recommendation condition is satisfied, improving the reliability and accuracy of the judgment of the function recommendation condition.

[0106] In some alternative implementations of the present disclosure, the determination of the initial function information of the target application in response to detecting that the function recommendation condition is met and based on the operation behavior data and historical manipulation data includes: in response to detecting that the function recommendation condition is met, extracting the user's user preference features based on the operation behavior data and historical manipulation data; matching the user preference features with the table preference features in the pre-constructed preference function matching table; in response to detecting that the user preference matches the table preference features, obtaining the function features of the table preference features from the preference function matching table; and determining the initial function information of the target application based on the function features.

[0107] In this alternative implementation, the user preference features are the preferences and habits demonstrated by the user when operating the target application, such as features in multiple aspects including the user's search behavior, click behavior, purchase behavior, evaluation behavior, etc.

[0108] In this alternative implementation, the function features refer to the specific functions and services provided by the target application, including the user interface, data processing, business logic, etc. The function part is the core of the application program and determines the main functions and characteristics of the target application. The preference function matching table is a pre-constructed corresponding table used to represent the corresponding relationship between the user's user preference features and the function features of the target application.

[0109] The method for determining the initial function information of the target application provided in this alternative implementation extracts the user's user preference features based on the operation behavior data and historical manipulation data; matches the user preference features with the table preference features in the pre-constructed preference function matching table; in response to detecting that the user preference matches the table preference features, obtains the function features of the table preference features from the preference function matching table; and determines the initial function information of the target application based on the function features, improving the reliability of the obtained initial function information.

[0110] In some alternative implementations of the present disclosure, the above operation behavior data includes: intention gesture data. The determination of the initial function information of the target application based on the operation behavior data and the user's historical manipulation data includes: determining the user's initial intention based on the intention gesture data; determining the historical action trend based on the user's historical manipulation data; determining the user's target intention based on the historical action trend and the user's initial intention, and taking the function information related to the user's target intention in the target application as the initial function information.

[0111] In this alternative implementation, the intention gesture data refers to the data including the user's intention and gesture actions. The intention gesture data helps the machine understand the user's intention and needs by capturing and analyzing the user's hand movements, thereby providing a more natural and intuitive interaction experience.

[0112] In this alternative implementation, the user's initial intention is the intention reflected by the operation behavior data, and the user's target intention is the final intention obtained by the execution entity through in-depth analysis of the user's actions.

[0113] In this alternative implementation, the historical action trend refers to the change trend of the user's various actions within a certain time range. The historical manipulation data includes historical behavior data and the functions corresponding to the historical behavior data. The historical action trend can be obtained by extracting the corresponding action data from the historical behavior data. The historical action trend reflects the action change state when the user operates the target application, and the user's initial intention combined with the user's historical action trend can better reflect the user's target intention.

[0114] In this alternative implementation, the characteristics of the terminal side (such as the screen of a mobile terminal) supporting multiple gesture operations can be utilized to analyze the combination of clicks and other gestures (such as swipes, long presses, pinches, etc.). Different gesture combinations may represent different intentions of the user. For example, the user first clicks on an icon, then long presses and swipes on the icon. According to this specific gesture combination, it is reflected that the user wants to use a certain function. Combining the user's historical manipulation data, if the user has not operated the operation sequence of this function in history, the execution entity can recommend related in-depth functions. For example, in an image editing application, this gesture combination may trigger advanced image adjustment functions, such as local contrast enhancement and image quality enhancement in picture preview.

[0115] The method for determining the initial function information of the target application provided by this alternative implementation determines the user's initial intention based on the intention gesture data, determines the historical action trend based on the user's historical manipulation data, determines the user's target intention based on the historical action trend and the user's initial intention, and uses the function information related to the user's target intention in the target application as the initial function information, improving the reliability of obtaining the initial function information.

[0116] In some alternative implementations of the present disclosure, the determination of the initial function information of the target application based on the operation behavior data and the user's historical manipulation data includes: extracting the action data and the number of times of the action data in a preset time period from the user's historical manipulation data and operation behavior data, calculating the operation frequency of the action data based on the preset time and the number of times; determining the user's emotional state and operation state based on the operation frequency; determining the function information corresponding to the action data based on the emotional state and the operation state, and using this function information as the initial function information.

[0117] In this alternative implementation, the action data is the data representing actions in the operation behavior data and the historical manipulation data, such as click actions, swipe actions, flip actions, etc.

[0118] In this alternative implementation, the preset time period can be a time period set based on development requirements. Dividing the number of action data by the time period equals the operation frequency, and the operation frequency reflects the user's operation rhythm. The emotional state refers to the emotional experience generated by the user in a specific situation, usually with strong subjective feelings and psychological reactions. The operation state refers to the focused state of the user when moving the action data.

[0119] In this alternative implementation, in addition to simply analyzing the action data, it also deeply studies the user's operation rhythm (such as rapid consecutive clicks, slow intermittent clicks, etc.). The user's emotional state (such as rapid clicking when excited, slow clicking when tired) and operation state (such as whether focused, whether in a hurry) are judged through the operation rhythm. For example, when it is detected that the user's clicking rhythm is rapid and frequent, combined with the user's operation habits, it is judged that the user is in an excited state with a strong desire for exploration. At this time, some challenging or stimulating functions are recommended, such as high-difficulty levels in games, extreme sports videos in video applications, etc.

[0120] The method for determining the initial function information of the target application provided by this alternative implementation extracts the action data and the number of action data in the preset time period from the user's historical manipulation data and operation behavior data, calculates the operation frequency of the action data based on the preset time and the number; determines the user's emotional state and operation state based on the operation frequency; determines the function information corresponding to the action data based on the emotional state and operation state, and uses this function information as the initial function information, improving the reliability of obtaining the initial function information.

[0121] Optionally, the above information optimization and recommendation method further includes: predicting the user's click behavior based on the operation behavior data and historical operation data, and preloading in advance the functions or content that may be clicked. For example, according to the user's historical operation habits and the current operation interface, the execution entity predicts that the user may click a certain button to enter a new page or function. Before the user actually clicks, relevant resources are preloaded in the background in advance, enabling a quick response when the user clicks and enhancing the user experience. For example, in a news application, it is predicted that the user may click on the details page of an article, and the article content and relevant pictures are preloaded in advance to reduce the user's waiting time.

[0122] In some alternative implementations of the present disclosure, the determination of the initial function information of the target application based on the operation behavior data and the user's historical manipulation data includes: determining the operation object identifier based on the operation behavior data; selecting the operation function related to the operation object identifier from the user's historical manipulation data, determining the function information of the operation function, and using this function information as the initial function information.

[0123] In this alternative implementation, the operation object identifier is a user identifier set for the user whose execution subject the information optimization recommendation method runs on and whose operation target is the application. Both the operation behavior data and the historical manipulation data include the operation object identifier, and the users in the operation behavior data and the historical manipulation data can be unified through the operation object identifier.

[0124] The method for determining the initial function information provided by this alternative implementation determines the operation object based on the operation behavior data, selects the operation functions related to the operation object from the user's historical manipulation data, determines the function information of the operation functions, and uses the function information as the initial function information, providing a reliable implementation method for obtaining the initial function information and improving the comprehensiveness of obtaining the initial function information.

[0125] Optionally, determining the initial function information of the target application based on the operation behavior data and the user's historical manipulation data includes: determining a sequence of operation object identifiers with the same function representation based on the operation behavior data; selecting the operation functions related to the sequence of operation object identifiers from the user's historical manipulation data, determining the function information of the operation functions, and using the function information as the initial function information. The operation object is a basic unit in the target application (such as a control, a component, etc.), and the basic unit in the target application can be determined through the operation object identifier. A series of operation objects in the target application correspond to one function, and the sequence of operation object identifiers also corresponds to one function.

[0126] In some alternative implementations of the present disclosure, the above initial function information includes an initial function identifier. Determining the initial operation sequence based on the initial function information and sending the initial prompt information of the initial operation sequence includes: determining a set of initial operation sequences based on the initial function identifier; selecting an initial operation sequence from the set of initial operation sequences and sending the initial prompt information of the initial operation sequence.

[0127] In this alternative implementation, each function in the target application has a corresponding function identifier, and the initial function information has an initial function identifier. Each function corresponds to one or more operation sequences, and at least one operation sequence corresponding to the initial function is the set of initial operation sequences. Selecting at least one operation sequence from the set of initial operation sequences as the initial operation sequence can improve the accuracy of the initial operation sequence. Selecting at least one operation sequence as the initial operation sequence can select a set number of operation sequences with a higher occurrence frequency in the set of initial operation sequences, or can also select a set number of operation sequences with a greater user preference degree in the set of initial operation sequences.

[0128] The method for determining the initial operation sequence provided by this alternative implementation determines the set of initial operation sequences based on the initial function identifier, selects the initial operation sequence from the set of initial operation sequences, and sends the initial prompt information of the initial operation sequence. Selecting the initial operation sequence from the initial operation sequences preset for the initial function improves the accuracy of the obtained initial operation sequence.

[0129] In some alternative implementations of the present disclosure, obtaining the recommended operation sequence based on the feedback data and the initial operation sequence includes: detecting, based on the feedback data, whether the user approves of the initial operation sequence; in response to detecting that the user approves of the initial operation sequence, using the initial operation sequence as the recommended operation sequence; in response to detecting that the user does not approve of the initial operation sequence, extracting the operation sequence with the highest operation frequency in the feedback data, and using this operation sequence and the initial operation sequence as the recommended operation sequence.

[0130] In this alternative implementation, the feedback data is the user's feedback on the initial operation sequence. When the feedback data indicates that the user agrees with the initial operation sequence, the user approves of the initial operation sequence; when the feedback data indicates that the user does not agree with the initial operation sequence, the user does not approve of the initial operation sequence. When the user does not approve of the initial operation sequence, the feedback data may include: the operation data of the operation sequence other than the initial operation sequence by the user, and determining the recommended operation sequence through the operation sequence other than the initial operation sequence in the feedback data.

[0131] The method for obtaining the recommended operation sequence provided by this alternative implementation detects, based on the feedback data, whether the user approves of the initial operation sequence. When the user approves of the initial operation sequence, using the initial operation sequence as the recommended operation sequence; when the user does not approve of the initial operation sequence, based on the operation sequence with the highest operation frequency in the feedback data, using this operation sequence and the initial operation sequence together as the recommended operation sequence, which improves the reliability of the obtained recommended operation sequence.

[0132] Optionally, obtaining the recommended operation sequence based on the feedback data and the initial operation sequence includes: detecting, based on the feedback data, whether the user approves of the initial operation sequence; in response to detecting that the user does not approve of the initial operation sequence, modifying the initial operation sequence based on the key information in the feedback data to obtain the recommended operation sequence.

[0133] In this embodiment, each operation sequence in the initial operation sequence has a weight corresponding to the degree of user interest. The feedback data includes: display feedback data, implicit feedback data, and rapid iteration mechanism data. The display feedback data includes the user manually adjusting the recommendation result. For example, for the display feedback data of "not interested", the above-mentioned modification of the initial operation sequence to obtain the recommended operation sequence includes: directly reducing the recommendation weight of the operation sequence corresponding to a certain function in the initial operation sequence, and re-ranking the initial operation sequence according to the weights to obtain the recommended operation sequence. Another example is the display feedback data of "recommend more relevant functions": the above-mentioned modification of the initial operation sequence to obtain the recommended operation sequence includes: recalculating the weights of the operation sequences in the initial operation sequence and sorting them to obtain the recommended operation sequence. The user can select to fix the commonly used functions to prevent some high-frequency functions from being ignored due to short-term non-use.

[0134] For the implicit feedback data, if the user ignores the recommended function multiple times, the system will automatically reduce the weight of the operation sequence corresponding to this function in the initial operation sequence. If the user uses the initial function, the system will increase the recommendation priority of the operation sequence of this function to enhance the personalized experience.

[0135] For the rapid iteration mechanism data, the user feedback can take effect immediately, and the execution entity will adjust the recommendation weight in real time to avoid recommendation failure or interference with the user experience. Combining with A / B testing, continuously optimize the recommendation algorithm to ensure the accuracy of the recommendation effect and user satisfaction.

[0136] Further referring to Figure 5 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of an information optimization recommendation device. This device embodiment corresponds to the Figure 1 method embodiment shown, and this device can be specifically implemented in various electronic devices.

[0137] As Figure 5 shown, the information optimization recommendation device 500 provided in this embodiment includes: a collection unit 501, a determination unit 502, a sending unit 503, a receiving unit 504, and a obtaining unit 505. Among them, the above-mentioned collection unit 501 can be configured to collect the operation behavior data of the user operating the target application. The above-mentioned determination unit 502 can be configured to determine the initial function information of the target application based on the operation behavior data and the user's historical manipulation data. The above-mentioned sending unit 503 can be configured to determine the initial operation sequence based on the initial function information and send the initial prompt information of the initial operation sequence. The above-mentioned receiving unit 504 can be configured to receive the feedback data of the user. The above-mentioned obtaining unit 505 can be configured to obtain the recommended operation sequence based on the feedback data and the initial operation sequence.

[0138] In this embodiment, in the information optimization recommendation device 500: For the specific processing of the acquisition unit 501, the determination unit 502, the sending unit 503, the receiving unit 504, and the obtaining unit 505 and the technical effects brought thereby, reference can be respectively made to Figure 1 the relevant descriptions of steps 101, 102, 103, 104, and 105 in the corresponding embodiment, which will not be elaborated here.

[0139] In some alternative implementation manners of this embodiment, the above determination unit 502 is configured to: Based on the operation behavior data and the user's historical manipulation data, detect whether the function recommendation condition is satisfied; In response to detecting that the function recommendation condition is satisfied, based on the operation behavior data and the historical manipulation data, determine the initial function information of the target application.

[0140] In some alternative implementation manners of this embodiment, the above determination unit 502 is configured to: Based on the operation behavior data and the user's historical manipulation data, determine the total operation times of the user for each function of the target application; Detect whether the total operation times is greater than or equal to the first set threshold; In response to detecting that the total operation times is greater than or equal to the first set threshold, determine that the function recommendation condition is satisfied.

[0141] In some alternative implementation manners of this embodiment, the above determination unit 502 is configured to: In response to detecting that the total operation times is less than the first set threshold, based on the operation behavior data and the historical manipulation data, determine the segmented operation times of this function in the time window; Detect whether the segmented operation times is greater than or equal to the second set threshold, and the second set threshold is less than the first set threshold; In response to detecting that the segmented operation times is greater than or equal to the second set threshold, determine that the function recommendation condition is satisfied.

[0142] In some alternative implementation manners of this embodiment, the above determination unit 502 is configured to: In response to detecting that the function recommendation condition is satisfied, input the operation behavior data and the historical manipulation data into the function recommendation model, and obtain the initial function information output by the function recommendation model, and the function recommendation model is installed on the terminal side.

[0143] In some alternative implementation manners of this embodiment, the above device further includes: an adjustment unit (not shown in the figure), and the above adjustment unit is configured to: Based on the feedback data, adjust the weights and parameters of the function recommendation model, and send the weights and parameters to the server.

[0144] In some alternative implementation manners of this embodiment, the above device further includes: an input unit (not shown in the figure), and the above input unit is configured to: Based on the historical manipulation data and the operation behavior data, determine the operation habit data of the user; Based on the operation habit data, determine the emotional state of the user; Input the emotional state into the function recommendation model.

[0145] In some alternative implementation manners of this embodiment, the determining unit 502 is configured to: determine user portrait data of a user based on operation behavior data and the user's historical manipulation data; detect whether the user portrait data matches the library portrait data of a corresponding function in a preset portrait database; in response to detecting that the user portrait data matches the library portrait data, determine that the function recommendation condition is satisfied.

[0146] In some alternative implementation manners of this embodiment, the determining unit 502 is configured to: extract user preference features of the user based on operation behavior data and historical manipulation data; match the user preference features with table preference features in a pre-constructed preference function matching table; in response to detecting that the user preference matches the table preference features, obtain function features of the table preference features from the preference function matching table; and determine initial function information of a target application based on the function features.

[0147] In some alternative implementation manners of this embodiment, the operation behavior data includes intent gesture data, and the determining unit 502 is configured to: determine an initial intent of the user based on the intent gesture data; determine a historical action trend based on the user's historical manipulation data; determine a target intent of the user based on the historical action trend and the user's initial intent, and use function information related to the user's target intent in the target application as the initial function information.

[0148] In some alternative implementation manners of this embodiment, the determining unit 502 is configured to: extract action data and the number of times of the action data in a preset time period from the user's historical manipulation data and operation behavior data, calculate an operation frequency of the action data based on the preset time and the number of times; determine an emotional state and an operation state of the user based on the operation frequency; determine function information corresponding to the action data based on the emotional state and the operation state, and use the function information as the initial function information.

[0149] In some alternative implementation manners of this embodiment, the determining unit 502 is configured to: determine an operation object identifier based on the operation behavior data; select an operation function related to the operation object identifier from the user's historical manipulation data, determine function information of the operation function, and use the function information as the initial function information.

[0150] In some alternative implementation manners of this embodiment, the initial function information includes an initial function identifier, and the sending unit 503 is configured to: determine an initial operation sequence set based on the initial function identifier; select an initial operation sequence from the initial operation sequence set, and send an initial prompt message of the initial operation sequence.

[0151] In some alternative implementation manners of this embodiment, the obtaining unit 505 is configured to: detect whether the user approves of the initial operation sequence based on the feedback data; in response to detecting that the user approves of the initial operation sequence, use the initial operation sequence as the recommended operation sequence; in response to detecting that the user does not approve of the initial operation sequence, extract the operation sequence with the highest operation frequency in the feedback data, and use this operation sequence and the initial operation sequence as the recommended operation sequence.

[0152] For the information optimization recommendation device provided in the embodiment of the present disclosure, first, the acquisition unit 501 acquires the operation behavior data of the user operating the target application; second, the determination unit 502 determines the initial function information of the target application based on the operation behavior data and the user's historical manipulation data; third, the sending unit 503 determines the initial operation sequence based on the initial function information, and sends the initial prompt information of the initial operation sequence; fourth, the receiving unit 504 receives the feedback data of the user; finally, the obtaining unit 505 obtains the recommended operation sequence based on the feedback data and the initial operation sequence. Thus, based on the user's operation behavior data on the terminal side, the initial operation sequence corresponding to the initial function information is determined, and the user can be effectively reminded of the functions that can be operated through the initial operation sequence, improving the user's function usage efficiency; based on the user's feedback data and the initial operation sequence, the recommended operation sequence is determined, improving the accuracy and reliability of the recommended operation sequence recommended to the user.

[0153] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information and other processes all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0154] According to the embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0155] Figure 6 The schematic block diagram of an exemplary electronic device 600 that can be used to implement the embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0156] As Figure 6As shown, device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to computer programs stored in a read-only memory (ROM) 602 or computer programs loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of device 600 can also be stored. The computing unit 601, ROM 602, and RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0157] Multiple components in device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a disk, optical disc, etc.; and a communication unit 609, such as a network card, modem, wireless communication transceiver, etc. The communication unit 609 allows device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0158] The computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 601 executes the various methods and processes described above, such as the information optimization recommendation method. For example, in some embodiments, the information optimization recommendation method can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as the storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed onto device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the information optimization recommendation method described above can be executed. Alternatively, in other embodiments, the computing unit 601 can be configured to execute the information optimization recommendation method by any other appropriate means (e.g., by means of firmware).

[0159] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0160] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable information optimization recommendation device, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowchart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0161] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0162] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, speech input, or tactile input).

[0163] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an information server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0164] A computer system can include a client and a server. The client and the server are generally far from each other and typically interact through a communication network. The client-server relationship is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, or a server of a distributed system, or a server incorporating a blockchain.

[0165] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitation is imposed herein.

[0166] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. An information optimization recommendation method, applied to the terminal side, the method comprising: Collecting operation behavior data of a user operating a target application; Determining initial function information of the target application based on the operation behavior data and the user's historical operation data; Determining an initial operation sequence based on the initial function information, and sending an initial prompt message of the initial operation sequence; Receiving feedback data of the user; Obtaining a recommended operation sequence based on the feedback data and the initial operation sequence.

2. The method according to claim 1, wherein The determining the initial function information of the target application based on the operation behavior data and the user's historical operation data includes: Detecting whether a function recommendation condition is satisfied based on the operation behavior data and the user's historical operation data; In response to detecting that the function recommendation condition is satisfied, determining the initial function information of the target application based on the operation behavior data and the historical operation data.

3. The method according to claim 2, wherein The detecting whether a function recommendation condition is satisfied based on the operation behavior data and the user's historical operation data includes: Determining the total number of operations of the user on each function of the target application based on the operation behavior data and the user's historical operation data; Detecting whether the total number of operations is greater than or equal to a first set threshold; In response to detecting that the total number of operations is greater than or equal to the first set threshold, determining that the function recommendation condition is satisfied.

4. The method according to claim 3, wherein The detecting whether a function recommendation condition is satisfied based on the operation behavior data and the user's historical operation data further includes: In response to detecting that the total number of operations is less than the first set threshold, determining the segmented number of operations of the function in a time window based on the operation behavior data and the historical operation data; Detecting whether the segmented number of operations is greater than or equal to a second set threshold, the second set threshold being less than the first set threshold; In response to detecting that the segmented number of operations is greater than or equal to the second set threshold, determining that the function recommendation condition is satisfied.

5. The method according to claim 2, wherein, The determining the initial function information of the target application based on the operation behavior data and the historical operation data in response to detecting that the function recommendation condition is satisfied includes: In response to detecting that the function recommendation condition is satisfied, inputting the operation behavior data and the historical operation data into a function recommendation model to obtain the initial function information output by the function recommendation model, the function recommendation model being installed on the terminal side.

6. According to the method of claim 5, the method further comprises: Adjusting the weights and parameters of the function recommendation model based on the feedback data, and sending the weights and parameters to the server.

7. According to the method of claim 5, the method further comprises: Determining the operation habit data of the user based on the historical operation data and the operation behavior data; Determining the emotional state of the user based on the operation habit data; Inputting the emotional state into the function recommendation model.

8. The method according to claim 2, wherein The detecting whether a function recommendation condition is satisfied based on the operation behavior data and the user's historical operation data includes: Determine the user portrait data of the user based on the operation behavior data and the historical manipulation data of the user; Detect whether the user portrait data matches the library portrait data of the corresponding function in the preset portrait database; In response to detecting that the user portrait data matches the library portrait data, determine that the function recommendation condition is satisfied.

9. The method according to claim 2, wherein The determining of the initial function information of the target application based on the operation behavior data and the historical manipulation data in response to detecting that the function recommendation condition is satisfied includes: In response to detecting that the function recommendation condition is satisfied, extract the user preference features of the user based on the operation behavior data and the historical manipulation data; Match the user preference features with the table preference features in the pre-constructed preference function matching table; In response to detecting that the user preference matches the table preference features, obtain the function features of the table preference features from the preference function matching table; Determine the initial function information of the target application based on the function features.

10. The method according to any one of claims 1-9, wherein, The operation behavior data includes: intention gesture data, and the determining of the initial function information of the target application based on the operation behavior data and the historical manipulation data of the user includes: Determine the initial intention of the user based on the intention gesture data; determine the historical action trend based on the historical manipulation data of the user; Determine the user target intention based on the historical action trend and the user initial intention, and use the function information related to the user target intention in the target application as the initial function information.

11. The method according to any one of claims 1-9, wherein, The determining of the initial function information of the target application based on the operation behavior data and the historical manipulation data of the user includes: Extract the action data and the number of times of the action data in a preset time period from the historical manipulation data of the user and the operation behavior data; Calculate the operation frequency of the action data based on the preset time and the number of times; Determine the emotional state and operation state of the user based on the operation frequency; Determine the function information corresponding to the action data based on the emotional state and the operation state, and use this function information as the initial function information.

12. The method according to any one of claims 1-9, wherein, The determining of the initial function information of the target application based on the operation behavior data and the historical manipulation data of the user includes: Determine the operation object identifier based on the operation behavior data; Select the operation function related to the operation object identifier from the historical manipulation data of the user, determine the function information of the operation function, and use this function information as the initial function information.

13. The method according to any one of claims 1-9, wherein, The initial function information includes an initial function identifier, and the determining of the initial operation sequence based on the initial function information and sending the initial prompt information of the initial operation sequence includes: Determine the initial operation sequence set based on the initial function identifier; Select an initial operation sequence from the initial operation sequence set and send the initial prompt information of the initial operation sequence.

14. The method according to any one of claims 1-9, wherein, The obtaining of the recommended operation sequence based on the feedback data and the initial operation sequence includes: Based on the feedback data, detect whether the user approves of the initial operation sequence; In response to detecting that the user approves the initial operation sequence, use the initial operation sequence as the recommended operation sequence; In response to detecting that the user does not approve the initial operation sequence, extract the operation sequence with the highest operation frequency in the feedback data, and use this operation sequence and the initial operation sequence as the recommended operation sequence.

15. An information optimization recommendation device, applied to the terminal side, the device includes: A collection unit, configured to collect operation behavior data of a user operating a target application; A determination unit, configured to determine initial function information of the target application based on the operation behavior data and the user's historical manipulation data; A sending unit, configured to determine an initial operation sequence based on the initial function information, and send an initial prompt message of the initial operation sequence; A receiving unit, configured to receive the feedback data of the user; A obtaining unit, configured to obtain a recommended operation sequence based on the feedback data and the initial operation sequence.

16. An electronic device, characterized in that, Includes: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1-14.

17. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the method according to any one of claims 1-14.

18. A computer program product, including a computer program, where the computer program implements the method according to any one of claims 1-14 when executed by a processor.

Images