Target feedback value determination method and apparatus, electronic device, and storage medium

Abstract

Embodiments of the application disclose a target feedback value determination method and device, electronic equipment and a storage medium. The method comprises: obtaining a sampling rate corresponding to each touch point, an object quantity and first data, the first data comprising target user data of a target user, first object data of a first object and online behavior data of the target user, the touch point being a contact point of the user and the object; determining a weight value of each touch point according to the sampling rate corresponding to each touch point and the object quantity; for each touch point, determining a first feedback value according to the target user data, the plurality of first object data and the online behavior data, the first feedback value being used to describe a positive feedback degree of the target user to the first object; and performing weighted calculation on the first feedback value corresponding to each touch point based on the weight value to obtain a target feedback value.

Description

Technical Field

[0001] This application belongs to the field of Internet technology, and in particular relates to a method, apparatus, electronic device and storage medium for determining target feedback value. Background Technology

[0002] With the development of information technology, big data and artificial intelligence are playing an increasingly important role in technological development. The application of data mining and intelligent recommendation can reduce the time users spend selecting products and other related needs, enabling users to quickly find what they need even in their busy schedules.

[0003] Traditional, simple data analysis and mining are no longer sufficient to meet the needs of production and daily life. There is a need to utilize certain technical means to deeply uncover the meaning behind the data. Therefore, a more accurate method is currently required to calculate the degree of positive feedback from users to an object. Summary of the Invention

[0004] This application provides a method, apparatus, device, and storage medium for determining target feedback values, which can solve the current problem of needing a more accurate way to calculate the degree of positive feedback from a user to an object.

[0005] In a first aspect, embodiments of this application provide a method for determining a target feedback value, the method comprising:

[0006] Acquire the sampling rate, number of objects, and first data for each touchpoint. The first data includes the target user data of the target user, the first object data of the first object, and the online behavior data of the target user. The touchpoint is the point of contact between the user and the object.

[0007] The weight value of each touch point is determined based on the sampling rate and the number of objects corresponding to each touch point;

[0008] For each touchpoint, a first feedback value is determined based on the target user data, multiple first object data, and online behavior data. The first feedback value is used to describe the degree of positive feedback from the target user to the first object.

[0009] The target feedback value is obtained by weighting the first feedback value corresponding to each touch point based on the weight value.

[0010] Secondly, embodiments of this application provide a target feedback value determination device, the target feedback value determination device comprising:

[0011] The acquisition module is used to acquire the sampling rate, number of objects and first data corresponding to each touchpoint. The first data includes the target user data of the target user, the first object data of the first object and the online behavior data of the target user. The touchpoint is the contact point between the user and the object.

[0012] The first determining module is used to determine the weight value of each touch point based on the sampling rate and the number of objects corresponding to each touch point;

[0013] The second determining module is used to determine a first feedback value for each touchpoint based on target user data, multiple first object data, and online behavior data. The first feedback value is used to describe the degree of positive feedback from the target user to the first object.

[0014] The weighting module is used to perform weighted calculations on the first feedback value corresponding to each touch point based on the weight value, so as to obtain the target feedback value.

[0015] Thirdly, embodiments of this application provide an electronic device, the device including: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the method as described in the first aspect or any possible implementation of the first aspect.

[0016] Fourthly, embodiments of this application provide a readable storage medium storing computer program instructions that, when executed by a processor, implement the method as described in the first aspect or any possible implementation thereof.

[0017] In this embodiment, by acquiring the sampling rate, number of objects, and first data corresponding to each touchpoint, the first data includes target user data of the target user, first object data of the first object, and online behavior data of the target user. The touchpoint is the point of contact between the user and the object. Based on the sampling rate and number of objects corresponding to each touchpoint, a weight value for each touchpoint is determined. This allows for the determination of the weight value for each touchpoint based on its specific characteristics, clarifying the influence of the first feedback value of each touchpoint on the target feedback value. Then, for each touchpoint, based on the target user data, multiple first object data, and online behavior data, a first feedback value describing the degree of positive feedback from the target user to the first object can be quickly and accurately determined. Finally, the first feedback values ​​corresponding to each touchpoint are weighted and calculated based on the weight values ​​to obtain the target feedback value. The target feedback value determined by this embodiment accurately expresses the degree of positive feedback from the target user to the first object. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method for determining a target feedback value provided in an embodiment of this application;

[0020] Figure 2 This is a schematic diagram of the structure of a target feedback value determination device provided in an embodiment of this application;

[0021] Figure 3 This is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0022] The features and exemplary embodiments of various aspects of this application will now be described in detail. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only configured to explain this application and are not configured to limit this application. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples of this application.

[0023] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, object, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, object, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, object, or apparatus that includes the element.

[0024] The target feedback value determination method provided in this application embodiment can be applied to at least the following application scenarios, which will be described below.

[0025] With the continuous development of information technology, focusing on users and deeply understanding their needs has become a key competitive advantage that operators urgently need to enhance. Currently, there is an urgent need to explore new market opportunities from the vast amount of information, find new business breakthroughs, and provide scientific business decision-making suggestions for business development.

[0026] "Data mining + intelligent recommendation" technology is an inevitable product of the big data era. With the rise of concepts such as intelligence and datafication, big data and artificial intelligence are playing an increasingly important role in technological development, being successfully applied to all aspects of life and work and continuously innovating. The application of "data mining + intelligent recommendation" technology can reduce the time users spend selecting products and other related needs, allowing them to quickly find what they need even in their busy schedules.

[0027] Traditional, simple data analysis and mining are no longer sufficient to meet the needs of production and daily life. Utilizing specific technologies to deeply uncover the meaning behind data and using the results of data analysis to guide scientific business deployment strategies is crucial. Data analysis and mining includes using mathematical methods to calculate relevant "feedback values" to achieve accurate recommendations and effective ranking for each user.

[0028] Figure 1 This is a flowchart of a target feedback value determination method provided in an embodiment of this application.

[0029] like Figure 1 As shown, the target feedback value determination method may include steps 110-140. This method is applied to the target feedback value determination device, as detailed below:

[0030] Step 110: Obtain the sampling rate, number of objects, and first data corresponding to each touchpoint. The first data includes the target user data of the target user, the first object data of the first object, and the online behavior data of the target user. The touchpoint is the contact point between the user and the object.

[0031] Step 120: Determine the weight value of each touch point based on the sampling rate and the number of objects corresponding to each touch point;

[0032] Step 130: For each touchpoint, determine the first feedback value based on the target user data, multiple first object data, and online behavior data. The first feedback value is used to describe the degree of positive feedback from the target user to the first object.

[0033] Step 140: Calculate the target feedback value by weighting the first feedback value corresponding to each touch point based on the weight value.

[0034] In this embodiment, by acquiring the sampling rate, number of objects, and first data corresponding to each touchpoint, the first data includes target user data of the target user, first object data of the first object, and online behavior data of the target user. The touchpoint is the point of contact between the user and the object. Based on the sampling rate and number of objects corresponding to each touchpoint, a weight value for each touchpoint is determined. This allows for the determination of the weight value for each touchpoint based on its specific characteristics, clarifying the influence of the first feedback value of each touchpoint on the target feedback value. Then, for each touchpoint, based on the target user data, multiple first object data, and online behavior data, a first feedback value describing the degree of positive feedback from the target user to the first object can be quickly and accurately determined. Finally, the first feedback values ​​corresponding to each touchpoint are weighted and calculated based on the weight values ​​to obtain the target feedback value. The target feedback value determined by this embodiment accurately expresses the degree of positive feedback from the target user to the first object.

[0035] The following describes the contents of steps 110-140 respectively:

[0036] Step 110 is involved.

[0037] Obtain the sampling rate, number of objects, and first data corresponding to each touchpoint. The first data includes the target user data, the first object data, and the target user's online behavior data. The touchpoint is the point of contact between the user and the object.

[0038] Typically, user and target touchpoints can include business websites, official accounts, social media platforms, and in-store customer interactions. Operators can use data analytics to integrate all these touchpoints and provide a unified and consistent experience for their customers. Data insights can be gained from multiple customer touchpoints, and analytics tools can effectively build brand awareness and improve customer satisfaction.

[0039] The sampling rate is the percentage of the number of users sampled at the current touchpoint relative to the total number of users.

[0040] For example, the collected samples include touchpoint 1 (app), touchpoint 2 (mini program), touchpoint 3 (wap hall), and touchpoint 4 (online hall). The sampling rate of touchpoint 1 is 0.6; the sampling rate of touchpoint 2 is 0.2; the sampling rate of touchpoint 3 is 0.1; and the sampling rate of touchpoint 4 is 0.1.

[0041] Promotional information on social media platforms can include H5 pages for marketing campaigns or offline posters, with QR codes on the posters. This information can be sent directly to users. After scanning the QR code, users can directly complete their transactions.

[0042] The target user data may include: gender, age, city, device, activity level, total monthly data usage, data usage within the plan, data usage outside the plan, monthly data usage within the plan, number of successful subscriptions, number of subscriptions only viewed, number of views of the target audience, number of successful activity subscriptions, number of clicks on data packages in the past two months, number of clicks on activities in the past two months, number of clicks on the target audience in the past two months, number of clicks on service-related items in the past two months, and number of activity shares in the past two months.

[0043] The target users' online behavior data includes: clicks, shares, and transactions related to objects. Clicks, shares, and transactions related to objects are rated 1, while no clicks, shares, or transactions are rated 0. Therefore, assuming M users and N objects, a rating matrix R = M * N can be constructed.

[0044] The first set of data for the first object includes: object business type, channel, click-through rate, click volume, pageviews, and impressions. Click volume comes from database activity records; each reported pageview record is confirmed as a click, and each reported impression is confirmed as an impression. Data package product characteristics include data package price, data package type, data package configuration validity period, number of data package purchases, and number of refunds.

[0045] Step 120 is involved.

[0046] The weight value of each touch point is determined based on the sampling rate and the number of objects corresponding to each touch point;

[0047] The formula for calculating the weight value can be:

[0048] Where M(i) represents the weight value corresponding to each touch point, 4 represents 4 preset touch points, α represents the sampling rate corresponding to each touch point, and β represents the number of objects corresponding to each touch point.

[0049] The collected samples include touchpoint 1 (app), touchpoint 2 (mini-program), touchpoint 3 (WAP hall), and touchpoint 4 (online hall), etc., and the collection rate of each touchpoint is calculated. The sampling rate is the proportion of the number of users sampled at the current touchpoint relative to the total number of users. The user ratio scheme of each touchpoint can be dynamically intervened by humans. The weight of the collection rate here will affect the dynamic weight adjustment of the feature coefficients in the feedback value calculation process.

[0050] Step 130 is involved.

[0051] For each touchpoint, a first feedback value is determined based on the target user data, multiple first object data, and online behavior data. The first feedback value is used to describe the degree of positive feedback from the target user to the first object.

[0052] In one possible embodiment, step 130 may specifically include the following steps:

[0053] Based on target user data, multiple first-object data, and online behavior data, determine the target user's second feedback value to the first object;

[0054] Based on the second feedback value, the target object is determined from the first object, and the second feedback value of the target user to the target object is greater than the first preset threshold.

[0055] The target user data and the target object data are input into a pre-trained prediction model, which outputs the first feedback value.

[0056] If the second feedback value of the target object is greater than the first preset threshold, then the candidate first object can be preliminarily screened, and the target object data of the screened target object is input into the pre-trained prediction model to output the first feedback value.

[0057] The system retrieves product recommendations to obtain objects for each user based on similarity matching. Each user and each object form a user feature table and an object feature table in the associated database. The corresponding feature values ​​are retrieved and subjected to certain feature transformations to obtain a suitable feature vector. The feature vector is then input into the "feedback value model" to calculate the ranking variable value. Finally, the ranking variable value is substituted into the activation function to obtain a continuous first feedback value in the range [0-1].

[0058] Before the target user data and target object data are input into the pre-trained prediction model and the first feedback value is output, the process also includes: building the prediction model.

[0059] First, prepare the training data: Associate each user and object with its feature table to obtain feature values. Expand and concatenate the user's and object's feature values ​​to form a feature vector. Then, perform feature processing on the training data, specifically dimensionality reduction, to obtain user feature sets and object feature sets.

[0060] Secondly, based on the processing of the feature data, a feature vector vector(u, i) is constructed, where each element in the vector represents a feature column. Assuming there are n elements, the n-dimensional model is designed as follows:

[0061] model=g(x)=w1x1+w2x2+w3x3+...+w n x n .

[0062] The selection of discrete classification features involves setting up experimental and control groups, controlling the discrete classification features as input variables, and observing whether these features affect the scoring results. For discrete classification features, assuming there are m categories, m state bits are set according to the number of categories (all state bits are 0 by default). The category being processed will be indicated by a value of 1 on the corresponding state bit, resulting in an m-bit two-dimensional array [0,1,0,0]. Feature vectors are constructed based on this array. Dimensionality reduction of the high-dimensional feature vectors involves calculating the eigenvalues ​​and eigenvectors of the covariance matrix. The eigenvectors are arranged into a matrix based on the eigenvalues ​​from largest to smallest, and the first k rows are used to form matrix p. Y = PX yields the dimensionality-reduced data Y.

[0063] Finally, the model is trained. Rating sampling involves randomly sampling 100 data points from an R = M*N rating matrix, each containing rating data for 50 objects. This is then transformed into 5000 data points. Each data point includes known rating information and its corresponding feature vector.

[0064] Based on the prediction model, the feature coefficients in the prediction model are obtained through machine learning iteration. At this time, the feature coefficients of the prediction model are established as (W1, W2, W3, W4, W5... Wn). These coefficients are then used to build the model.

[0065] Each feature vector in the dataset is input into the model to obtain the observed value f(x). After processing by the model, the observed value f(x) is compared with the true value y. If they are the same, the prediction is correct; otherwise, the prediction is incorrect.

[0066] Where N = 5000, f(xi) represents the observed value of the i-th data point calculated by the model, and yi represents the original true score value carried by the i-th data point. The calculation results indicate the accuracy of the model to some extent. The smaller the root mean square error, the better the simulation effect; the larger the error, the worse the effect.

[0067] For example, feature = (1, 42, 551, 13542, 23555, 0.05, 2, 8...); After three steps—discrete classification feature selection, two-dimensional encoding of classification features, and high-order dimensionality reduction projection—we obtain: feature = (1, 42, 1, 0, 0, 0, 0.325, 0.458, 0.05, 2, 8, 0, 0, 0, 1...). Substituting these values ​​into the model, we calculate the model value: model = g(x) = w1x1 + w2x2 + w3x3 + ... + w n x n The final prediction model was derived.

[0068] Here, a model for the feedback value algorithm is established, and a large amount of data from the database is input into the algorithm model. As the user data continues to increase, the machine learning cycle time continues to lengthen, and the accuracy of the iterative data continues to improve, the feature coefficients of the algorithm model are finally calculated: (W1, W2, W3, W4, W5... Wn).

[0069] Specifically, the step of determining the second feedback value of the target user to the first object based on the target user data, multiple first object data, and online behavior data may include the following steps:

[0070] Acquire online behavior data of multiple first users, where the feedback value of the first user to the first object is greater than a second preset threshold;

[0071] Based on the online behavior data of the first user and the online behavior data of the target user, determine the first similarity between the target user and the first user;

[0072] Based on the first similarity and the first user's feedback value to the first object, determine the target user's second feedback value to the first object.

[0073] Based on the online behavior data of the first user and the target user, the first similarity between the target user and the first user is determined, which can be calculated using the following formula:

[0074]

[0075] xi represents the target user's actions such as liking, forwarding, and purchasing the first object; yi represents the first user's actions such as liking, forwarding, and purchasing the first object.

[0076] For example, the similarity value between user a and user b is 0.27; and so on, a similarity-based relationship network matrix between users is obtained, that is, the first similarity between the target user and the first user is determined:

[0077]

[0078] Finally, the second feedback value of the target user to the first object, as mentioned above, is determined based on the first similarity and the first user's feedback value to the first object. This can be calculated using the following formula:

[0079]

[0080] Where p(u, i) represents the degree of interest that the target user u has in each first object i, i.e., the second feedback value;

[0081] Set S(u,k): represents the k first users most similar to the target user u.

[0082] Set N(advert): Represents all users who are interested in the first object, advert.

[0083] w uv Let S(u,k) represent the similarity between user u and each similar user in user set v, where user set v is the intersection of set S(u,k) and set N(advert).

[0084] r vi This represents the degree to which user v likes the first object i (feedback value). i can be advertise1, advertise2, etc.

[0085] For example, to recommend target objects to target user u, select K=3 similar users. Based on the relationship network matrix, find the similar users: u2, u3, and u4. Then, the objects they are interested in but u is not interested in are: sim1(u2, u4) and sim2(u3, u4). Calculate p(u, advert1) and p(u, advert2) respectively. Summarize the object preferences for each user based on the above calculation results to obtain the recommended content.

[0086] Specifically, the steps involved in obtaining online behavior data from multiple first users may include the following:

[0087] Obtain second user data from multiple second users;

[0088] Based on the target user data and the second user data, determine the second similarity between the target user and the second user;

[0089] Based on the second similarity, at least one first user is identified from multiple second users, and online behavior data of multiple first users is obtained.

[0090] First, based on the target user data and the second user data, determine the second similarity between the target user and the second user, and then calculate the similarity value between target user a and second user b.

[0091] Where xi represents the target user's actions such as liking, forwarding, and purchasing the first object; yi represents the second user's actions such as liking, forwarding, and purchasing the first object.

[0092] By analogy, a similarity-based relationship network matrix among users is obtained. Then, from the user relationship network matrix, the K users most similar to the target user u are found. That is, based on the second similarity, at least one first user is determined from multiple second users, denoted by the set S(u, K).

[0093] Step 140 is involved.

[0094] The target feedback value is obtained by weighting the first feedback value corresponding to each touch point based on the weight value.

[0095] Providing users with recommended objects can help them choose objects and services, and provide assistance for user decision-making. It has an unparalleled advantage in Internet applications. The target feedback value is obtained by weighting the first feedback value corresponding to each touchpoint based on the weight value.

[0096] Therefore, based on the target feedback value, when users do not have a clear target, it can help them discover new information and new needs that are of interest, and provide personalized services, which are more intelligent and proactive.

[0097] In one possible embodiment, step 140 may specifically include the following steps:

[0098] When the target user interacts with a third object, obtain the adjustment value corresponding to the third object;

[0099] The third feedback value is obtained by weighting the first feedback value corresponding to each touch point based on the weight value.

[0100] The target feedback value is determined based on the adjustment value and the third feedback value.

[0101] The target feedback value reflects the distance between the currently associated object and the target user. In the process of calculating the target feedback value, certain special user indicators must also be considered. These special user indicators are reflected through a third object. When there is interaction between the target user and the third object, the adjustment value pt corresponding to the third object is obtained.

[0102] The first feedback value corresponding to each contact point: model = g(x) = w1x1 + w2x2 + w3x3 + ... + w n x n ;

[0103] Weight values: Where α is the sampling rate and β is the number of product objects;

[0104] The first feedback value corresponding to each touch point is weighted and calculated based on the weight value to obtain the third feedback value, which is M(i)*f(g(x));

[0105] The adjustment value is pt, which reflects the distance between the currently associated object and the target user;

[0106] The target feedback value is determined based on the adjustment value and the third feedback value: that is, the target feedback value score = M(i)*f(g(x))+Pt.

[0107] Currently, among all the products recommended by a user, each product is sorted from largest to smallest based on the calculated feedback value, and the results are stored in the database.

[0108] Specifically, the steps mentioned above regarding obtaining the adjustment value corresponding to the third object when there is interaction between the target user and the third object may include the following steps:

[0109] When there is interaction between the target user and a third object, obtain the third object data of the third object;

[0110] The adjustment value is determined based on the data from the third object.

[0111] The adjustment value reflects the alienation between the currently associated object and the target user. In the process of calculating the target feedback value, certain special user indicators should also be considered. These indicators are usually not statistically significant, but rather come from strategic planning and business orientation. These special user indicators need to be reflected through the adjustment value.

[0112] For example, if a special promotional campaign for a traffic package is designed in a capsule position within a client application, then more attention will be paid to the user's historical data on traffic usage.

[0113] The third object includes, but is not limited to, "last month's usage within the package", "last month's usage outside the package", "this month's data renewal amount", "number of check-ins", and "task completion rate".

[0114] Adjustment value This reflects the adjustment and control of the feedback score. When a user performs well in certain specific metrics, a higher pt value is given, and vice versa.

[0115] λ represents the baseline value of the nth indicator, Rsi represents the error value of the feedback value model, and Pt plays a prominent role in the calculation of the feedback value for indicators specified by specific business orientations.

[0116] Specifically, the step of weighting the first feedback value corresponding to each touch point based on the weight value to obtain the third feedback value may include the following steps:

[0117] The activation function is used to calculate the activation of the first feedback value corresponding to each touch point to obtain the fourth feedback value.

[0118] The third feedback value is obtained by weighting the fourth feedback value corresponding to each touch point based on the weight value.

[0119] Activation function: The symbol e is a constant in mathematics, an infinite non-repeating decimal, and a transcendental number, with a value of approximately 2.718.

[0120] First feedback value: model = g(x) = w1x1 + w2x2 + w3x3 + ... + w n x n ;

[0121] Based on the activation function, the activation calculation is performed on the first feedback value corresponding to each touch point to obtain the fourth feedback value: f(g(x));

[0122] The fourth feedback value corresponding to each touch point is calculated by weighting the weight value to obtain the third feedback value: M(i)*f(g(x)); where M(i) is the weight value.

[0123] The target feedback value is then obtained as score = M(i)*f(g(x)) + Pt.

[0124] The target feedback value determination method provided in this application obtains the sampling rate, number of objects, and first data corresponding to each touchpoint. The first data includes target user data of the target user, first object data of the first object, and online behavior data of the target user. The touchpoint is the contact point between the user and the object. Based on the sampling rate and number of objects corresponding to each touchpoint, a weight value for each touchpoint is determined. This allows for the determination of the weight value for each touchpoint based on its specific characteristics, clarifying the influence of the first feedback value of each touchpoint on the target feedback value. Then, for each touchpoint, based on the target user data, multiple first object data, and online behavior data, a first feedback value describing the degree of positive feedback from the target user to the first object can be quickly and accurately determined. Finally, the first feedback values ​​corresponding to each touchpoint are weighted and calculated based on the weight values ​​to obtain the target feedback value. The target feedback value determined through the embodiments of this application can accurately express the degree of positive feedback from the target user to the first object.

[0125] Based on the above Figure 1 The method for determining the target feedback value shown in the present application also provides a device for determining the target feedback value, such as... Figure 2 As shown, the target feedback value determination device 200 may include:

[0126] The acquisition module 210 is used to acquire the sampling rate, number of objects and first data corresponding to each touchpoint. The first data includes the target user data of the target user, the first object data of the first object and the online behavior data of the target user. The touchpoint is the contact point between the user and the object.

[0127] The first determining module 220 is used to determine the weight value of each touch point according to the sampling rate and the number of objects corresponding to each touch point;

[0128] The second determining module 230 is used to determine a first feedback value for each touchpoint based on target user data, multiple first object data and online behavior data. The first feedback value is used to describe the degree of positive feedback from the target user to the first object.

[0129] The weighting module 240 is used to perform weighted calculations on the first feedback value corresponding to each touch point based on the weight value to obtain the target feedback value.

[0130] In one possible embodiment, the second determining module 230 is specifically used for:

[0131] Based on the target user data, multiple first object data, and the online behavior data, determine the second feedback value of the target user to the first object;

[0132] Based on the second feedback value, a target object is determined from the first object, wherein the second feedback value of the target user to the target object is greater than a first preset threshold;

[0133] The target user data and the target object data are input into a pre-trained prediction model, and the first feedback value is output.

[0134] In one possible embodiment, the second determining module 230 is specifically used for:

[0135] Acquire online behavior data of multiple first users, where the feedback value of the first user to the first object is greater than a second preset threshold;

[0136] Based on the online behavior data of the first user and the online behavior data of the target user, a first similarity between the target user and the first user is determined;

[0137] Based on the first similarity and the first user's feedback value to the first object, the second feedback value of the target user to the first object is determined.

[0138] In one possible embodiment, the second determining module 230 is specifically used for:

[0139] Obtain second user data from multiple second users;

[0140] Based on the target user data and the second user data, a second similarity between the target user and the second user is determined;

[0141] Based on the second similarity, at least one first user is determined from the plurality of second users, and online behavior data of the plurality of first users is obtained.

[0142] In one possible embodiment, the weighting module 240 is specifically used for:

[0143] When the target user interacts with a third object, the adjustment value corresponding to the third object is obtained;

[0144] Based on the weight values, the first feedback value corresponding to each of the touch points is weighted and calculated to obtain the third feedback value;

[0145] The target feedback value is determined based on the adjustment value and the third feedback value.

[0146] In one possible embodiment, the weighting module 240 is specifically used for:

[0147] When the target user interacts with a third object, the third object data of the third object is obtained;

[0148] The adjustment value is determined based on the data of the third object.

[0149] In one possible embodiment, the weighting module 240 is specifically used for:

[0150] Based on the activation function, the activation calculation is performed on the first feedback value corresponding to each of the touch points to obtain the fourth feedback value;

[0151] The third feedback value is obtained by weighting the fourth feedback value corresponding to each touch point based on the weight value.

[0152] In this embodiment, by acquiring the sampling rate, number of objects, and first data corresponding to each touchpoint, the first data includes target user data of the target user, first object data of the first object, and online behavior data of the target user. The touchpoint is the point of contact between the user and the object. Based on the sampling rate and number of objects corresponding to each touchpoint, a weight value for each touchpoint is determined. This allows for the determination of the weight value for each touchpoint based on its specific characteristics, clarifying the influence of the first feedback value of each touchpoint on the target feedback value. Then, for each touchpoint, based on the target user data, multiple first object data, and online behavior data, a first feedback value describing the degree of positive feedback from the target user to the first object can be quickly and accurately determined. Finally, the first feedback values ​​corresponding to each touchpoint are weighted and calculated based on the weight values ​​to obtain the target feedback value. The target feedback value determined by this embodiment accurately expresses the degree of positive feedback from the target user to the first object.

[0153] Figure 3 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of this application is shown.

[0154] An electronic device may include a processor 301 and a memory 302 storing computer program instructions.

[0155] Specifically, the processor 301 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0156] Memory 302 may include mass storage for data or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 302 may include removable or non-removable (or fixed) media. Where appropriate, memory 302 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 302 is non-volatile solid-state memory. In a particular embodiment, memory 302 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0157] The processor 301 reads and executes computer program instructions stored in the memory 302 to implement any of the target feedback value determination methods in the embodiment shown in the figure.

[0158] In one example, the electronic device may also include a communication interface 303 and a bus 310. For example, Figure 3 As shown, the processor 301, memory 302, and communication interface 303 are connected through bus 310 and complete communication with each other.

[0159] The communication interface 303 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0160] Bus 310 includes hardware, software, or both, that couples components of an electronic device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 310 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0161] The electronic device can execute the target feedback value determination method in the embodiments of this application, thereby achieving a combination Figure 1 The method for determining the target feedback value is described.

[0162] Furthermore, in conjunction with the target feedback value determination method in the above embodiments, this application embodiment can provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; these computer program instructions are implemented when executed by a processor. Figure 1 Method for determining the target feedback value.

[0163] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0164] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0165] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0166] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for determining a target feedback value, characterized in that, The method includes: Acquire the sampling rate, number of objects, and first data corresponding to each touchpoint. The first data includes target user data of the target user, first object data of the first object, and online behavior data of the target user. The touchpoint is the contact point between the user and the object. The weight value of each touch point is determined based on the sampling rate and the number of objects corresponding to each touch point; For each of the touchpoints, a first feedback value is determined based on the target user data, multiple first object data, and the online behavior data. The first feedback value is used to describe the degree of positive feedback from the target user to the first object. The target feedback value is obtained by weighting the first feedback value corresponding to each touch point based on the weight value. For each of the touchpoints, a first feedback value is determined based on the target user data, multiple sets of first object data, and the online behavior data, including: Based on the target user data, multiple first object data, and the online behavior data, determine the second feedback value of the target user to the first object; Based on the second feedback value, a target object is determined from the first object, wherein the second feedback value of the target user to the target object is greater than a first preset threshold; The target user data and the target object data of the target object are input into a pre-trained prediction model, and the first feedback value is output. The step of determining the second feedback value of the target user to the first object based on the target user data, multiple sets of first object data, and the online behavior data includes: Acquire online behavior data of multiple first users, where the feedback value of the first user to the first object is greater than a second preset threshold; Based on the online behavior data of the first user and the online behavior data of the target user, a first similarity between the target user and the first user is determined; Based on the first similarity and the first user's feedback value to the first object, the second feedback value of the target user to the first object is determined.

2. The method according to claim 1, characterized in that, The acquisition of online behavior data from multiple first users includes: Obtain second user data from multiple second users; Based on the target user data and the second user data, a second similarity between the target user and the second user is determined; Based on the second similarity, at least one first user is determined from the plurality of second users, and online behavior data of the plurality of first users is obtained.

3. The method according to claim 1, characterized in that, The step of weighting the first feedback value corresponding to each touch point based on the weight value to obtain the target feedback value includes: When the target user interacts with a third object, the adjustment value corresponding to the third object is obtained; Based on the weight values, the first feedback value corresponding to each of the touch points is weighted and calculated to obtain the third feedback value; The target feedback value is determined based on the adjustment value and the third feedback value.

4. The method according to claim 3, characterized in that, When the target user interacts with a third object, obtaining the adjustment value corresponding to the third object includes: When the target user interacts with a third object, the third object data of the third object is obtained; The adjustment value is determined based on the data of the third object.

5. The method according to claim 3, characterized in that, The step of weighting the first feedback value corresponding to each touch point based on the weight value to obtain the third feedback value includes: Based on the activation function, the activation calculation is performed on the first feedback value corresponding to each of the touch points to obtain the fourth feedback value; The third feedback value is obtained by weighting the fourth feedback value corresponding to each touch point based on the weight value.

6. A target feedback value determination device, characterized in that, The target feedback value determination device includes: The acquisition module is used to acquire the sampling rate, number of objects and first data corresponding to each touchpoint. The first data includes the target user data of the target user, the first object data of the first object and the online behavior data of the target user. The touchpoint is the contact point between the user and the object. The first determining module is used to determine the weight value of each touch point according to the sampling rate and the number of objects corresponding to each touch point; The second determining module is used to determine a first feedback value for each of the touchpoints based on the target user data, multiple first object data, and the online behavior data, wherein the first feedback value is used to describe the degree of positive feedback from the target user to the first object; The weighting module is used to perform weighted calculation on the first feedback value corresponding to each of the touch points based on the weight value to obtain the target feedback value; The second determining module includes: Based on the target user data, multiple first object data, and the online behavior data, determine the second feedback value of the target user to the first object; Based on the second feedback value, a target object is determined from the first object, wherein the second feedback value of the target user to the target object is greater than a first preset threshold; The target user data and the target object data of the target object are input into a pre-trained prediction model, and the first feedback value is output. The step of determining the second feedback value of the target user to the first object based on the target user data, multiple sets of first object data, and the online behavior data includes: Acquire online behavior data of multiple first users, where the feedback value of the first user to the first object is greater than a second preset threshold; Based on the online behavior data of the first user and the online behavior data of the target user, a first similarity between the target user and the first user is determined; Based on the first similarity and the first user's feedback value to the first object, the second feedback value of the target user to the first object is determined.

7. An electronic device, characterized in that, The electronic device includes: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the target feedback value determination method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, implement the target feedback value determination method as described in any one of claims 1-5.

Images