CN114118651A - Evaluation method, device, equipment and computer storage medium - Google Patents

Abstract

The application provides an evaluation method, an evaluation device, evaluation equipment and a computer storage medium, relates to the technical field of computers, and aims to improve the accuracy of evaluation on an object producer. The method comprises the following steps: obtaining object evaluation values of at least two target objects of a target object producer to be evaluated, wherein the object evaluation value of each target object is determined according to behavior characteristic values of multiple dimensions of a historical user, and the behavior characteristic values of the multiple dimensions are determined according to behavior data generated by the historical user aiming at the target object; and performing target fusion processing on the object evaluation values of the at least two target objects to obtain an evaluation result of a target object producer, wherein the target fusion processing comprises an operation of improving the influence degree of the target object of which the object evaluation value meets the business target on the evaluation result. According to the method, the evaluation result of the target object producer is obtained based on the object evaluation value of the target object created by the target object producer, and the accuracy of evaluating the target object producer is improved.

Description

Evaluation method, device, equipment and computer storage medium

Technical Field

The present application relates to the field of computer technologies, and in particular, to an evaluation method, an evaluation device, evaluation equipment, and a computer storage medium.

Background

When evaluating an object producer authoring an object in a target sharing platform, the related art evaluates the object producer authoring the object according to feedback information of a user of the pushed object on the object producer, but in the method for evaluating the object producer, the evaluation result of the object is not associated, so that the evaluation result of the object producer is inaccurate, and therefore, how to improve the accuracy of evaluating the object producer becomes a problem to be considered.

Disclosure of Invention

The embodiment of the application provides an evaluation method, an evaluation device, evaluation equipment and a computer storage medium, which are used for improving the accuracy of evaluation on an object producer.

In a first aspect of the present application, an evaluation method is provided, including:

obtaining object evaluation values of at least two target objects related to a target object producer to be evaluated, wherein the object evaluation value of each target object is determined according to behavior characteristic values of multiple dimensions of a historical user, and the behavior characteristic values of the multiple dimensions are determined according to behavior data of the historical user aiming at the target object;

and performing target fusion processing on the object evaluation values of the at least two target objects to obtain an evaluation result of the target object producer, wherein the target fusion processing comprises an operation of improving the influence degree of the target object with the object evaluation value meeting the business target on the evaluation result.

In a possible implementation manner, the object browsing characteristic value includes at least one of the following browsing index values:

a browsing completion index value obtained by processing a browsing completion rate of the target object, wherein the browsing completion rate is determined based on a ratio of the browsing completion times of the target object to the total browsing times;

a browsing duration index value which is determined according to the duration information of the target object browsed by the historical user;

a browsing number index value determined by a total number of times the target object is browsed by the historical user;

the browsing rate index value is determined based on the browsing rate of the target object, and the browsing rate is the ratio of the number of the historical users browsing the target object to the total number of the historical users corresponding to the target object.

In a second aspect of the present application, there is provided an evaluation apparatus, comprising:

the evaluation system comprises a first evaluation unit, a second evaluation unit and a third evaluation unit, wherein the first evaluation unit is used for obtaining object evaluation values of at least two object objects related to a target object producer to be evaluated, the object evaluation value of each object is determined according to behavior characteristic values of multiple dimensions of a historical user, and the behavior characteristic values of the multiple dimensions are determined according to behavior data generated by the historical user aiming at the object;

and the second evaluation unit is used for performing target fusion processing on the object evaluation values of the at least two target objects to obtain an evaluation result of the target object producer, and the target fusion processing comprises operation of improving the influence degree of the target object of which the object evaluation value meets the business target on the evaluation result.

In a possible implementation manner, the second evaluation unit is specifically configured to:

weighting the object evaluation value of each of the at least two target objects by using a preset evaluation reference value to obtain a weighted object evaluation value of each target object, so that the first evaluation value deviation degree of the at least two target objects is higher than the second evaluation value deviation degree; the first evaluation value deviation degree is determined according to the deviation degree of the weighted object evaluation values of the first part of object objects and the weighted object evaluation values of the second part of object objects, and the second evaluation value deviation degree is determined according to the deviation degree of the object evaluation values of the first part of object objects and the object evaluation values of the second part of object objects; the first part of target objects comprise target objects of which the object evaluation values meet the business target in the at least two target objects, and the second part of target objects comprise target objects except the first part of target objects in the at least two target objects;

and fusing the weighted object evaluation reference values of the target objects to obtain the evaluation result of the target object producer.

In a possible implementation manner, the second evaluation unit is specifically configured to:

determining the evaluation reference value as a power, determining the power of the object evaluation value of each target object as a weighted object evaluation value of each target object, wherein the evaluation reference value is greater than 1;

the fusing the weighted object evaluation reference values of the target objects to obtain the evaluation result of the target object producer comprises the following steps:

determining the evaluation reference value as a root index, and performing evolution processing on the sum of the weighted object evaluation values of all the target objects to obtain an object comprehensive evaluation value;

and obtaining the evaluation score of the target object producer based on the object comprehensive evaluation value.

In a possible implementation manner, the second evaluation unit is specifically configured to:

obtaining an authoring stability parameter of the target object producer by using a preset confidence coefficient parameter and the total number of the target objects based on the following formula, wherein the authoring stability parameter represents the stability of the target object producer in authoring the number of the target objects meeting the business objective:

in the formula, the R is an authoring stability parameter of the target object producer; the above-mentioned

e is a natural base number; the N is the total number of the target objects; the CL is the confidence

A parameter; the q is a preset confidence coefficient reference value;

and adjusting the comprehensive evaluation value of the object by using the creation stability parameter to obtain the evaluation score of the evaluation target object producer.

In a possible implementation manner, the target objects whose object evaluation values meet the business target include the at least two target objects, and after the target objects are sorted in the order from high to low according to the object evaluation values, the target objects which are sorted in the front are sorted; or

The target object with the object evaluation value meeting the business target comprises at least two target objects, and the object evaluation value meets the target object of the object evaluation threshold.

In a possible implementation manner, the first evaluation unit is specifically configured to:

for any one of the at least two target objects:

determining behavior characteristic values of multiple dimensions of the target object according to behavior data, generated by a historical user corresponding to the target object, for the target object;

determining the importance of the behavior characteristic value of each dimension in the behavior characteristic values of the plurality of dimensions based on the influence value between the behavior characteristic values of every two dimensions in the behavior characteristic values of the plurality of dimensions;

and according to the importance of the behavior characteristic value of each dimension, carrying out weighting processing on the importance of the behavior characteristic value of each dimension to obtain an object evaluation value of the target object.

In one possible implementation, the behavior feature values of the plurality of dimensions include at least two feature values:

an object interaction feature value; the object interaction characteristic value determines data of interactive operation on the target object according to the historical user;

an object browsing characteristic value; the object browsing characteristic value is determined according to the data of the target object browsed by the historical user;

a resource permutation eigenvalue; and determining the resource replacement characteristic value according to the historical user and aiming at the data of the target object transferred electronic resource.

In a possible implementation manner, the object browsing characteristic value includes at least one of the following index values:

a browsing completion index value obtained by processing a browsing completion rate of the target object, wherein the browsing completion rate is determined based on a ratio of the browsing completion times of the target object to the total browsing times;

a browsing duration index value which is determined according to the duration information of the target object browsed by the historical user;

a browsing number index value determined by a total number of times the target object is browsed by the historical user;

the browsing rate index value is determined based on the browsing rate of the target object, and the browsing rate is the ratio of the number of the historical users browsing the target object to the total number of the historical users corresponding to the target object.

In a third aspect of the present application, there is provided a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of the first aspect and any one of the possible embodiments when executing the program.

In a fourth aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program comprising computer instructions, the computer instructions being stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method provided in the various possible implementations of the first aspect described above.

In a fifth aspect of the present application, there is provided a computer-readable storage medium having stored thereon computer instructions which, when executed on a computer, cause the computer to perform the method of any one of the first aspect and any one of the possible embodiments.

Due to the adoption of the technical scheme, the embodiment of the application has at least the following technical effects:

in the embodiment of the application, the object evaluation values of a plurality of object objects created by an object producer are subjected to object fusion processing to obtain the evaluation result of the object producer, on one hand, the object evaluation values of the object objects are determined based on behavior characteristic values of a plurality of dimensions of a historical user, namely, the object evaluation values reflect the preference degree of the historical user to the object objects, and further, the evaluation result of the object producer obtained by performing object fusion on the object evaluation values reflects the preference degree of the historical user to the object objects created by the object producer; on the other hand, in the target fusion processing, the influence of the target object with the object evaluation value meeting the business target on the evaluation result is improved, so that the evaluation result of the target object producer with the created target object meeting the business target is obviously different from that of other target producers, and the accuracy of evaluating the target producer is improved.

Drawings

Fig. 1 is an exemplary diagram of an application scenario provided in an embodiment of the present application;

FIG. 2 is a schematic flow chart diagram of an evaluation method according to an embodiment of the present disclosure;

FIG. 3 is a schematic flowchart illustrating a specific example of an evaluation method according to an embodiment of the present application;

fig. 4 is a schematic diagram of an architecture of an evaluation principle provided in an embodiment of the present application

Fig. 5 is a schematic structural diagram of an evaluation device according to an embodiment of the present application;

fig. 6 is a schematic structural diagram of a computer device according to an embodiment of the present application.

Detailed Description

In order to better understand the technical solutions provided by the embodiments of the present application, the following detailed description is made with reference to the drawings and specific embodiments.

In order to facilitate those skilled in the art to better understand the technical solutions of the present application, the following description refers to the technical terms of the present application.

Object and object producer: the object can be content or an article created on the target sharing platform, the content can be but is not limited to pictures and texts, videos, audios, game props in game applications, software applications and the like, the article can be but is not limited to food, clothes, stationery and the like, and a person skilled in the art can set the object according to actual needs; the object producer refers to an individual or a group which creates the object, that is, an object producer may refer to a user, or may refer to a user who creates the object belonging to a group, and those skilled in the art can determine the object producer according to actual needs.

Target object and target object producer: in the embodiment of the application, the target object producer refers to an object producer to be evaluated, and the target object refers to an object created by the object producer to be evaluated.

A target sharing platform: a platform for authoring and sharing objects; the target sharing platform generally comprises an object creation producer, a platform operation and an object consumer (namely, a user browsing the object); the target sharing platform in the embodiment of the application may include, but is not limited to, a content sharing platform or an article replacement platform, and the content sharing platform may include, but is not limited to, an image-text platform (such as an information platform and a social platform), a video platform (such as a short video platform and a live broadcast platform), an audio platform (such as a music platform and a book listening platform), and the like.

And (4) service target: the target sharing platform is used for sharing the operation target of the target sharing platform, the service target is not limited too much in the embodiment of the application, and a person skilled in the art can set the operation target according to actual service requirements; for example, but not limited to, the business target may be set as an object producer that determines a high-quality object authored in the target sharing platform, the business target may also be set as an object producer that determines the most popular with consumers on the target sharing platform, or the business target may be set as an object producer that finds an object having a potential to author a high-quality object in the target sharing platform, and the like.

The following explains the concept of the present application.

When an object producer and an object of an object created in a target sharing platform are evaluated, the evaluation of the object is scored through the attribute characteristics of the object in the process of evaluating a single object in the related technology; in the process of evaluating a single object producer, the object producer who creates the object is often evaluated according to the feedback information of the user of the pushed object to the object producer; if the target sharing platform is a content sharing platform, evaluating and scoring the content (the object) through the attribute characteristics of the content, and evaluating the creator (the object producer) by the feedback information of the user of the pushed content on the creator creating the content; however, in the above evaluation process, the evaluation process of the object and the evaluation process of the object producer are performed separately, there is no correlation between the two, and the evaluation process of the object producer and the evaluation result of the object created by the object producer have no clear correlation, and the correlation degree between the evaluation result of the object producer and the evaluation result of the object created by the object producer cannot be directly reflected, so that the evaluation accuracy of the object producer is affected, and therefore how to improve the evaluation accuracy of the object producer becomes a problem to be considered.

In view of this, the inventor designs an evaluation method, an evaluation device, an evaluation apparatus, and a computer storage medium, considering that in a target sharing platform, a service target of the target sharing platform is considered, and an evaluation process of an object producer on the target sharing platform has a strong association relationship with an object created by the object producer on the target sharing platform, in this embodiment of the present application, each target object of a plurality of target objects created by the object producer to be evaluated is evaluated first to obtain an object evaluation value of each target object, and then the object evaluation values of the plurality of target objects created by the object producer are subjected to target fusion processing to obtain an evaluation result of the object producer; considering that the influence degrees of a plurality of target objects of an object producer on an evaluation target are different, in order to improve the influence degree of a high-value object with a high object evaluation value, the target fusion processing mainly comprises the operation of improving the influence degree of the target object with the object evaluation value meeting a service target on the evaluation result of the object producer; considering that the evaluation accuracy of the target object created by the object producer has a direct relationship with the evaluation accuracy of the object producer, the evaluation accuracy of the object producer is further improved by improving the evaluation accuracy of the target object in the embodiment of the application, and specifically, when the target object is evaluated, the object evaluation value of the target object is obtained through behavior characteristic values of multiple dimensions of a history user corresponding to the target object, so as to improve the accuracy of the object evaluation value; wherein the behavior characteristic values of the multiple dimensions are determined according to behavior data generated by historical users aiming at the target object.

In order to more clearly understand the design idea of the present application, the following provides an example description of an application scenario related to the embodiments of the present application.

Referring to fig. 1, a schematic structural diagram of a system 100 of a target sharing platform is provided, where the system of the target sharing platform includes at least one terminal device 110 and a server 120, where the terminal device 110 and the server 120 may communicate with each other through a network, and the communication network may be, but is not limited to, a local area network, a wide area network, and the like.

An object producer can log in a related interface of the target sharing platform through a terminal device 110 (such as but not limited to 110-1 or 110-2 in the figure), and then an object is authored through the related interface; the terminal device 110 transmits the object authored by the object producer to the server 120; the user can receive and browse the object pushed by the server 120 through the terminal device 110; the user is an object consumer, the object producer and the object consumer in the embodiment of the application can be the same individual or group, and the roles of the object producer and the object consumer can be interchanged under different actual scenes.

The server 120 (such as but not limited to the server 120-1, 120-2, or 120-3 in the figure) determines behavior feature values of multiple dimensions of each target object for the historical user according to behavior data generated by the historical user for each target object authored by the target object producer, determines an object evaluation value of each target object based on the behavior feature values of the multiple dimensions, and performs target fusion processing on the object evaluation values of the target objects to obtain an evaluation result of the target object producer.

The server 120 may be a server in the target sharing platform or a server outside the target sharing platform; when the server 120 is a server in the target sharing platform, the server 120 may receive an object sent by the terminal device 110, and push the object to the terminal device 110 according to the related information;

when the server 120 is a server outside the target sharing platform, the server 120 may obtain behavior data generated by the history user for each target object created by the target object producer from the server in the target sharing platform, perform an evaluation process according to the obtained behavior data, and may also send an evaluation result of the target object producer to the server in the target sharing platform; as described herein by taking the server 120-1 in fig. 1 as a server in the target sharing platform, and the server 120-2 as a server for implementing the evaluation process outside the target sharing platform as an example, specifically, the server 120-2 may be selected from the server 120-1, acquiring behavior data generated by historical users aiming at each target object authored by a target object producer, determining behavior characteristic values of the historical user for multiple dimensions of each target object according to the acquired behavior data, and determining an object evaluation value of each target object based on the behavior feature values of the plurality of dimensions, then the object evaluation value of the target object is subjected to target fusion processing to obtain the evaluation result of the target object producer, and the server 120-2 may also send the evaluation result of the target object producer to the server 120-1.

In the embodiment of the present application, the terminal device 110 may be, but is not limited to, an electronic device used by the user, and the electronic device may be a computer device having a certain computing capability and running instant messaging software and websites or social networking software and websites, such as a personal computer, a mobile phone, a tablet computer, a notebook, an e-book reader, and the like. Each terminal device 110 is connected to the server 120 through a wireless Network, and the server 120 may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a CDN (Content Delivery Network), and a big data and artificial intelligence platform.

Based on the application scenario in fig. 1, the following exemplarily illustrates an evaluation method in the embodiment of the present application, which is applied to the server 120, please refer to fig. 2, and the evaluation method in the embodiment of the present application includes the following steps:

step S201, obtaining object evaluation values of at least two target objects associated with a target object producer to be evaluated, where the object evaluation value of each target object is determined according to behavior feature values of multiple dimensions of a historical user, and the behavior feature values of the multiple dimensions are determined according to behavior data generated by the historical user for the target object.

Specifically, for any one of the at least two target objects, in order to improve the accuracy of the object evaluation value of the target object, but not limited to, the behavior feature values of multiple dimensions of the target object may be determined according to the behavior data generated by the historical user corresponding to the target object for the target object, and the object evaluation value of the target object may be determined based on the behavior feature values of the multiple dimensions.

It should be noted that, in this embodiment of the present application, behavior feature values of multiple dimensions of a target object are determined according to behavior data generated by a historical user corresponding to the target object in a target time period, where the historical user corresponding to the target object is a user corresponding to a terminal device that a server pushes the target object in the target time period; the target time period may be any time period before the current time, and those skilled in the art may set the target time period according to actual requirements, for example, setting the time from the current time to a week before the current time, or a month before the current time to the current time; in this embodiment of the present application, multiple dimensions of the behavior feature value of the target object may be, but are not limited to, an object interaction dimension, an object browsing dimension, a resource replacement dimension, and the like, and a person skilled in the art may also set other dimensions of the behavior feature value according to the business objective, such as a dimension of a user who brings new attention to an object producer from the object, and the like.

Step S202, carrying out target fusion processing on the object evaluation values of the at least two target objects to obtain the evaluation result of a target object producer; and the target fusion processing comprises the operation of improving the influence degree of the target object with the object evaluation value meeting the business target on the evaluation result.

The business target in the embodiment of the application is not limited too much, and a person skilled in the art can set the business target according to actual business requirements; if the business target is set as an object producer for determining the high-quality object created in the target sharing platform, the high-quality object can be determined by the object evaluation value of the object, and the higher the object evaluation value is, the better the quality of the object is; when the business target is set to determine a potential object producer in the target sharing platform, the potential object producer may be an object producer whose probability of creating a high-quality object satisfies a probability threshold, and the like.

The target object whose object evaluation value meets the business target in the embodiment of the application can be, but is not limited to, a high-quality object (i.e., an object with high popularity of the user), and the high-quality object can be a target object which is ranked in front after being ranked according to the order of the object evaluation values from high to low in at least two target objects created by a target object producer; if but not limited to, the target objects ranked in the order of the object evaluation values from high to low and ranked at the positions of the previous set proportion are determined as the good-quality objects, the set proportion can be but is not limited to 80%, 90% and 95%; the target object with the object evaluation value satisfying the object evaluation threshold value may also be determined as the high-quality object, where a ratio of the object evaluation threshold value to a preset maximum value of the object evaluation value satisfies a target proportion, for example, when the preset maximum value of the object evaluation value is set to 100, and the target proportion is 80%, the object evaluation threshold value may be set to any value between 80 and 100, but not limited thereto.

As an example, in the above step S201, considering that, when an object is evaluated, weights of behavior feature values of multiple dimensions of a target object are set manually in a normal condition, but there are correlation relationships among the behavior feature values of multiple dimensions of one target object, and these correlation relationships may affect an object evaluation value of the target object, the weights of the behavior feature values of the various dimensions are set manually, and there is a great deviation with a mutual influence degree between the behavior feature values of the various dimensions, which further causes the weights of the behavior feature values of the various dimensions set manually to be inaccurate, and affects accuracy of an evaluation result of the object, so in order to further improve accuracy of the object evaluation value of the target object, in the embodiment of the present application, for one target object, based on influence values between behavior feature values of every two dimensions of the behavior feature values of the multiple dimensions of the target object, determining the importance of the behavior characteristic value of each dimension in the behavior characteristic values of the plurality of dimensions, and performing weighting processing on the importance of the behavior characteristic value of each dimension according to the importance of the behavior characteristic value of each dimension to obtain an object evaluation value of the target object.

Specifically, the weight of the behavior characteristic value of each dimension can be determined through an Analytic Hierarchy Process (AHP), the behavior characteristic values of multiple dimensions are compared pairwise to determine a proper scale, the pairwise comparison result of the behavior characteristic values of different dimensions is filled into a pairwise comparison matrix to obtain a value in the pairwise comparison matrix, then the expert gives the set weight of the behavior characteristic value of each dimension, and the importance of the behavior characteristic value of each dimension is obtained through verification of the characteristic value.

E.g. the value a of the result of comparing the behavior characteristic value i of one dimension of the plurality of dimensions with the behavior characteristic value j of another dimension_ijFilling the comparison matrix into the ith row and the jth column of the paired comparison matrix, filling the pairwise comparison results of the behavior characteristic values of multiple dimensions into the paired matrix to obtain the paired comparison matrix, wherein the behavior characteristic value i of one dimension is compared with the behavior characteristic value j of the other dimension to obtain a comparison result a_ijThe values of (a) include the values in table 1 below.

Table 1:

comparing the behavior characteristic value i of one dimension with the behavior characteristic value j of another dimension	Score value
		Of equal importance	1
Of slight importance	3
		Of greater importance	5
Of strong importance	7
		Of extreme importance	9
Intermediate values of two adjacent judgments	2、4、6、8

In table 1, the left column indicates the result of comparing the behavior feature value i with the behavior feature value j, and the right column indicates the value a of the result of comparing the behavior feature value i with the behavior feature value j_ij(ii) a Where the behavior characteristic value i and the behavior characteristic value j are equally important, a_ijIs 1; when the behavior characteristic value i is slightly more important than the behavior characteristic value j, a_ijIs 3; when the behavior characteristic value i is more important than the behavior characteristic value j, a_ijIs 5; when the behavior characteristic value i is more important than the behavior characteristic value j, a_ijIs 7; when the behavior characteristic value i is extremely important than the behavior characteristic value j, a_ijIs 9; when the result of comparing the behavior characteristic value i with the behavior characteristic value j is the intermediate value of two adjacent results of the above equal importance, slight importance, strong importance and extreme importance, a_ijThe value of (d) may correspond to 2, 4, 6 or 8.

In the embodiment of the present application, the weight of the behavior feature value of each dimension in the behavior feature values of multiple dimensions of one target object may also be determined through an Attention Mechanism (Attention Mechanism), which is not described herein too much, and a person skilled in the art may set a specific implementation manner according to actual needs.

The following describes the behavior feature value of the target object according to the embodiment of the present application in detail.

In this embodiment of the present application, the behavior feature values of multiple dimensions of one target object include one or more feature values of an object interaction feature value, an object browsing feature value, and a resource replacement feature value, where:

the object interaction characteristic value is determined according to data of interactive operation of a historical user on the target object; the object interaction characteristic value is determined according to data of interactive operation on the target object by a user corresponding to the terminal equipment which pushes the target object in the target time period; the interaction operation may include, but is not limited to, a like operation for the target object, a comment operation for commenting on the target object in a corresponding area of the target object, and a sharing operation for sharing the target object with other users in the form of a sharing code (e.g., a web link or a two-dimensional code).

The object interaction characteristic value of the target object can be determined by the principle of the following formula 1.

Score_com＝w1×Rate_like+w2×Rate_comment+w3×Rate_shareFormula (1)

In equation 1, Score_comAn object interaction characteristic value of a target object; rate_likeThe target object is an approval rate of the object, which may be, but is not limited to, a ratio of the number of times that the target object is subjected to the approval operation by the history user to the number of times that the target object is browsed by the history user; rate_commentThe comment rate of the object can be but is not limited to the ratio of the number of times that the target object is subjected to comment operations of the historical user to the number of times that the target object is browsed by the historical user; rate_shareThe target object sharing rate is the ratio of the number of times of sharing operation of the historical user on the target object to the number of times of browsing the target object by the historical user; w1, w2 and w3 are the weights of the approval operation, the comment operation and the sharing operation respectively.

W1, w2, and w3 in formula 1 may be determined by the number of history users who perform each interactive operation (i.e., the like operation, the comment operation, and the share operation) on the target object, specifically, if the number of people who perform a certain interactive operation (e.g., the share operation) is small, and it is considered that the cost of performing the interactive operation by the user is higher, a higher weight is given to the interactive operation, and if the number of history users who perform the like operation is N1, the number of history users who perform the comment operation is N2, and the number of history users who perform the share operation is N3, w 1: w 2: w3 is set to N3: n2: n1.

W1, w2, and w3 in formula 1 may also be determined by the total number of times of each interactive operation (i.e., the approval operation, the review operation, and the sharing operation) for the target object, specifically, if the total number of times of a certain interactive operation (e.g., the sharing operation) is small, it may be considered that the cost of the user performing the interactive operation is higher than the cost of performing other interactive operations, and a higher weight is given to the interactive operation, if the total number of the approval operations is M1, the total number of the review operations is M2, and the total number of the sharing operations is M3, w 1: w 2: w3 was set at M3: m2: m1.

Those skilled in the art can also set the weights of the approval operation, the comment operation and the share operation in formula 1 in other ways based on the actual business objective.

The object browsing characteristic value in the embodiment of the application can be determined according to data of a target object browsed by a historical user; that is, the object browsing characteristic value may be determined according to the data of the target object browsed by the user corresponding to the terminal device that is pushed with the target object in the target time period, where the data of the target object browsed may include, but is not limited to, a browsing duration, a browsing start time, a browsing end time, a browsing frequency, and the like.

The resource replacement characteristic value in the embodiment of the application can be determined according to historical users and aiming at data of electronic resource transferred by a target object; the resource replacement characteristic value can be determined according to the data of the electronic resource transferred by the target object by the user corresponding to the terminal device which is pushed to the target object in the target time period; wherein transferring the electronic resource may include, but is not limited to, a payment operation for transferring electronic money for the target object, a gift operation for gifting an electronic gift for the target object, and the like.

The resource replacement characteristic value may be a total resource amount or an average resource amount of the electronic resource transferred by the historical user for the target object, or a total number of times or an average number of times of transferring the electronic resource by the historical user for the target object, and the like.

Further, the object browsing characteristic value in the embodiment of the present application includes at least one index value of a browsing completion index value, a browsing duration index value, a browsing number index value, and a browsing rate index value, and specific determination manners of the index values are described below.

The browsing completion index value is obtained by processing the browsing completion rate of the target object, and the browsing completion rate is determined based on the ratio of the browsing completion times of the target object to the total browsing times;

the browsing completion index value of a target object can be obtained through the following formula 2.

Score_f＝wilson_score(Rate_finish) Formula (2)

In equation 2, Score_fA browsing index value of a target object; rate_finishThe browsing completion rate of the target object is the ratio of the browsing duration of the target object to the physical duration of the target object; wison _ score () is a data check method.

The browsing completion rate of the target object in the embodiment of the application is obtained by the following method:

for a target object, determining the browsing duration of the target object browsed each time, further determining the browsing completion degree value of the target object browsed each time according to the browsing duration of the target object and the physical duration of the target object during each browsing, and further determining the average value of the browsing completion degree values of the target object browsed each time as the browsing completion rate of the target object; that is, the browsing completion rate of the target object is determined based on the following formula 3.

In equation 3, Rate_finishIs the browsing completion rate of a target object; i is an identifier of browsing times of the target object when being browsed by the historical user; rate_finish-iIs the browsing completion degree value of the target object browsed the ith time, and VV is the total browsing times of the target object by the history user.

When the target object is a video, the browsing duration is the duration of the video actually played by the historical user, and the physical duration is the duration required for playing the video; when the target object is audio, the browsing duration is the duration of the target object actually played by the historical user, and the physical duration is the duration required for playing the audio; if the target object is a video, the physical time length of the video is 100 minutes, and the time length for actually playing the video at a certain time is 80 minutes, the browsing completion degree value of the current browsing of the video is 80/100; if the video is browsed 5 times, the browsing completion degree value of each browsing is 80/100, 20/100, 40/100, 10/100 and 10/100, respectively, and the browsing completion rate of the video is (80/100+20/100+40/100+10/100+10/100)/5, that is, the browsing completion rate of the video is 32/100.

Since the browsing completion rate of the target object is affected by the total number vv of times that the target object is browsed by the historical user, a random error may exist, which affects the browsing completion rate of the target object, and further affects the accuracy of the browsing completion index value of the target object, in the embodiment of the present application, the winson _ score () is introduced to reduce the random error, and specifically, the principle of reducing the random error by the winson _ score () may be referred to the following formula 4.

In formula 4, p is the browsing completion Rate of a target object (i.e., the Rate in formula 2 and formula 3 above)_finish) (ii) a VV is the total number of times that the target object is browsed by the historical user; alpha is 0.005 and alpha is,

is 2.58²。

In the embodiment of the application, the index value of the browsing duration can be determined according to the duration information of the target object browsed by the historical user; specifically, the index value of the browsing duration of one target object may be obtained by, but is not limited to, the principle illustrated by the following formula 5.

In equation 5, Score_durationIs a browsing duration index value of a target object; duration is the target pairLike total duration of browsing in the current target period; duration_midThe target sharing platform is a browsing duration reference value which is a median of total browsing durations of all objects in a last target time period on the target sharing platform; browsing duration function value for naming a target object

Then Score is determined_95％F (duration) corresponding to all target objects in the last target time period on the target sharing platform are arranged in a descending order and then are sorted at the position of 95%;

wherein:

f(duration)_95％is Score in the above equation 5_95％(ii) a The above equation 5 can be modified to the following equation 6.

In equation 6, Score_durationIs a browsing duration index value of a target object; f (duration) is a browsing duration function value of the target object in the current target period; f (duration)_95％In the last target period on the target sharing platform, f (duration) corresponding to all target objects are arranged in descending order and then sorted to the 95% position.

The index value of the number of browsing times is determined by the total number of times the target object is browsed by the history user, and specifically, but not limited to, the index value of the number of browsing times of one target object may be obtained by the following formula 7.

In equation 7, Score_VVIs a target object's index value of the number of browsing times; VV is the target object at the current targetTotal number of views in a segment; VV_midA reference value of browsing times, which is a median of total browsing times of all objects (including objects authored by target object producers and objects authored by object producers other than the target object producer) in a last target time period on the target sharing platform; browsing times function value for naming a target object

Then Score is determined_95％In the last target time period on the target sharing platform, f (vv) corresponding to all target objects are arranged in descending order and then sorted at 95% of positions.

Wherein:

f(VV)_95％is Score in the above equation 7_95％The above equation 7 can be modified to the following equation 8.

In equation 8, Score_VVIs a target object's index value of the number of browsing times; f (vv) is a function value of the browsing times of the target object in the current target period; f (VV)_95％In the last target time period on the target sharing platform, f (vv) corresponding to all target objects are arranged in descending order and then sorted at 95% of positions.

The browsing rate index value is determined based on a browsing rate of a target object, wherein the browsing rate is a ratio of the number of historical users browsing the target object to the total number of the historical users corresponding to the target object; specifically, the browsing rate index value of one target object may be obtained by, but is not limited to, the following formula 9.

Score_str＝wilson_score(Rate_ctr) Formula (9)

In equation 9, Score_strIs a target object's browsing rate indicatorA value; rate_ctrThe browsing probability of the target object is the ratio of the number of times that the target object is browsed by the historical user in the target time period to the number of times that the target object is pushed.

As an embodiment, in step S202, a preset evaluation reference value may be used to perform weighting processing on an object evaluation value of each of at least two object created by an object producer, so as to obtain a weighted object evaluation value of each object; and performing fusion processing on the weighted object evaluation values of the target objects to obtain an evaluation result of a target object producer, wherein:

the purpose of the weighting process is to increase the degree of deviation between a first part of target objects and a second part of target objects, wherein the first part of target objects can be good quality objects, that is, the first part of target objects can be target objects satisfying business objectives, and the second part of target objects comprises target objects except the first part of target objects in the at least two target objects; the weighting process may make a first evaluation value deviation degree of the at least two target objects higher than a second evaluation value deviation degree; the first evaluation value deviation degree is determined according to the deviation degree of the weighted object evaluation values of the first part of the object objects and the weighted object evaluation values of the second part of the object objects, and the second evaluation value deviation degree is determined according to the deviation degree of the weighted object evaluation values of the first part of the object objects and the weighted object evaluation values of the second part of the object objects.

In the embodiment of the present application, the first evaluation value deviation and the second evaluation value deviation may be determined by, but not limited to, the following methods:

determining a first average value of the weighted object evaluation values of the first part of the target objects, determining a second average value of the weighted object evaluation values of the second part of the target objects, and determining a difference value between the first average value and the second average value as a first evaluation value deviation.

Determining a third average value of the pre-weighted object evaluation values of the first part of target objects, determining a fourth average value of the pre-weighted object evaluation values of the second part of target objects, and determining a difference value between the third average value and the fourth average value as a second evaluation value deviation.

Further, in the embodiment of the present application, the weighted object evaluation value of each of the at least two target objects may be obtained by, but is not limited to, the following weighting processing manner.

The first weighting processing method: and weighting the object evaluation values of the first part of target objects by using a preset evaluation reference value.

Specifically, the object evaluation values of the first part of target objects may be weighted by using a first evaluation reference value; if the first evaluation reference value is greater than 1, the first part of target objects may be weighted based on the following formula 10a to obtain weighted object evaluation values of the target objects in the first part of target objects, and the object evaluation values of the target objects in the second part of target objects are used as weighted object evaluation values thereof; if the first evaluation reference value is a value greater than 0 and less than 1, the first part of the target objects may be weighted based on the following formula 10b to obtain weighted object evaluation values of the target objects in the first part of the target objects, and the object evaluation values of the target objects in the second part of the target objects may be used as the weighted object evaluation values.

w_Score_i＝Score_iXrefer 1 formula (10a)

In equations 10a and 10b, i is the identification of each target object in the first part of target objects; w _ Score_iIs the weighted object evaluation value of the target object identified as i; score_iIs the object evaluation value before weighting of the target object identified as i; refer1 is the first evaluation reference value described above.

Or weighting the object evaluation values of the first part of target objects by using a preset second evaluation reference value; the second evaluation reference value may be a value smaller than the minimum value of the object evaluation values of the first part of the target objects, and if the minimum value of the object evaluation values of the first part is 90, the second evaluation reference value may be, but is not limited to, set to 75 or 80, etc.; further, it is possible, but not limited to, to perform weighting processing on the object evaluation values of the respective object objects in the first part of the object objects by using the second evaluation reference value based on the following formula 10c to obtain weighted object evaluation values of the respective object objects in the first part of the object objects, and to use the object evaluation values of the respective object objects in the second part of the object objects as the weighted object evaluation values thereof.

In formula 10c, i is the identifier of each target object in the first part of target objects; w _ Score_iIs the weighted object evaluation value of the target object identified as i; score_iIs the object evaluation value before weighting of the target object identified as i; the refscore 2 is the second evaluation reference value described above.

The second weighting processing method: and weighting the object evaluation values of the second part of target objects by using preset evaluation reference values.

Specifically, the third evaluation reference value may be used to perform weighting processing on the object evaluation values of the second part of target objects; if the third evaluation reference value is greater than 1, the second part of target objects may be weighted based on the following formula 11a to obtain weighted object evaluation values of the target objects in the second part of target objects, and the object evaluation values of the target objects in the first part of target objects are used as weighted object evaluation values thereof; if the third evaluation reference value is a value greater than 0 and less than 1, the second part of the target objects may be weighted based on the following formula 11b to obtain weighted object evaluation values of the target objects in the second part of the target objects, and the object evaluation values of the target objects in the first part of the target objects may be used as the weighted object evaluation values.

w_Score_j＝Score_jXrefer 3 formula (11b)

In equations 11a and 11b, j is the identifier of each target object in the second part of target objects; w _ Score_jIs the weighted object evaluation value for the target object identified as j; score_jIs the object evaluation value before weighting of the target object identified as i; refer3 is the third evaluation reference value described above.

Or weighting the object evaluation values of the second part of target objects by using a preset fourth evaluation reference value; the fourth evaluation reference value may be a value greater than the maximum value of the object evaluation values of the second part of the target objects, and if the maximum value of the object evaluation values of the second part of the target objects is 40, the second evaluation reference value may be, but is not limited to, set to 50 or 60, etc.; further, it is possible, but not limited to, to perform weighting processing on the object evaluation values of the respective object in the second part of the object objects by using the second evaluation reference value based on the following formula 11c to obtain weighted object evaluation values of the respective object in the second part of the object objects, and to use the object evaluation values of the respective object in the first part of the object objects as the weighted object evaluation values thereof.

In formula 11c, j is the identifier of each target object in the second part of target objects; w _ Score_jIs the weighted object evaluation value for the target object identified as j; score_jIs the object evaluation value before weighting of the target object identified as j; the refscore 4 is the fourth evaluation reference value described above.

The third weighting processing mode: and weighting the object evaluation values of the first part of target objects and the second part of target objects by using preset evaluation reference values.

Specifically, the preset fifth evaluation reference value may be determined as a power based on the following formula 12, and the power of the object evaluation value of each of the first part of the target objects and the second part of the target objects may be determined as each of the first part of the target objects and the second part of the target objects; the fifth evaluation reference value is a value greater than 1, the value of the fifth evaluation reference value is not limited too much in the embodiment of the present application, and a person skilled in the art can set the fifth evaluation reference value according to actual requirements, for example, the fifth evaluation reference value is set to 10 or 8;

in formula 12, i is the identity of each target object; w _ Score_iIs a weighted object evaluation value for the target object identified as i; score_iIs the object evaluation value before weighting of the target object identified as i; p is the above-mentioned fifth evaluation reference value.

As an embodiment, after the third weighting processing method is used to obtain the object evaluation reference value of each of the at least two target objects created by the target object producer in step S202, the fifth evaluation reference value may be, but is not limited to, determined as a root index, and the sum of the weighted object evaluation values of each target object is subjected to the evolution processing to obtain the object comprehensive evaluation value; obtaining an evaluation score of a target object producer based on the object comprehensive evaluation value; such as using the above-mentioned object comprehensive evaluation value as the evaluation score of the target object producer, etc.

That is, the evaluation reference value may be used as a norm, and the target comprehensive evaluation value may be obtained based on the following formula 13.

In formula 13, Score is an object comprehensive evaluation value of the target object producer; p is the above-mentioned fifthEvaluating a reference value, i identification of a target object authored by a target object producer, Score_iN is the total number of at least two target objects authored by the target object producer for the object evaluation value of the target object identified as i.

As an embodiment, considering the stability of the target object producer creating the high-quality object, the stability is strongly related to the evaluation of the target object producer, therefore, in the above step S202, based on the following formula 14, the authoring stability parameter of the target object producer is obtained by using the preset confidence parameter and the total number of the target objects created by the target object producer in the target time period, the authoring stability parameter represents the stability of the number of the target objects that the target object producer creates and meets the above business goal, and the comprehensive evaluation value of the objects is adjusted by using the authoring stability parameter, so as to obtain the evaluation score of the target object producer.

In formula 14, R is an authoring stability parameter of the evaluation target object producer; e is a natural base number; n is the total number of the target objects; the CL is the confidence coefficient parameter; q is a preset confidence reference value, and in the embodiment of the present application, q is 1, and CL is 95% for an exemplary description, a person skilled in the art may set q and CL according to actual requirements.

Specifically, the product of the authoring stability parameter and the object composite score may be determined as the evaluation score of the target object producer based on the following formula 15.

In equation 15, Score_authorIs an evaluation score of a target object producer; 1 is a preset confidence reference value; CL is a confidence parameter; i is the target objectAn identification of a target object created by the producer; score_iIs an object evaluation value for the target object identified as i; score_i ^pIs a weighted object evaluation value for the target object identified as i; p is an evaluation reference value; n is the number of target objects authored by the target object producer.

As an embodiment, after step S202, according to the evaluation score of the target object producer, a producer rating of the target object producer may be determined, or a ranking list may be performed on the target object producer, or a potential producer may be determined, which is further described below.

In the embodiment of the application, the object producers with evaluation scores in different score intervals can be divided into the object producers with different grades, for example, 5 producer grades such as excellent, good, common, poor and poor are set, the interval consisting of the preset highest value and the preset lowest value of the evaluation scores of the object producers is sequentially divided into 5 continuous score intervals, and the divided 5 continuous score intervals are respectively matched with the set 5 producer grades one by one according to the sequence of the evaluation scores from high to low; further, after the evaluation score of the target object producer is determined in step S202, the producer grade matching the score interval corresponding to the evaluation score is determined as the producer grade of the target object producer; it should be noted that, in the embodiment of the present application, the number of the divided score intervals, the corresponding relationship between each score interval and the producer grade, and the setting of the producer grade are not limited, and a person skilled in the art may set the producer grade and the matching relationship between the producer grade and the score interval according to actual needs; wherein the producer rating may reflect the popularity of the subject producer.

After the step S201 and the step S202 are performed, the evaluation scores of a plurality of object producers in the target sharing platform can be determined, and then the object producers can be ranked according to the order of the evaluation scores from high to low or from low to high, so as to obtain the producer ranking list.

After the evaluation score of the target object producer is determined in the step S202, if the evaluation score of the target object producer is greater than the evaluation score threshold of the potential producer, determining the target object producer as the potential object producer of the target sharing platform; a person skilled in the art can set the evaluation score threshold of the potential producer according to actual requirements, for example, when the preset maximum value of the evaluation score of the object producer in the target sharing platform is 100, the evaluation score threshold of the potential producer can be set to 70, 75, 80, or the like; or the object producer with the evaluation score increment larger in the last target period can be determined as the potential producer.

A specific example of evaluating an object producer is given below.

In this example, the content sharing platform is used as the target sharing platform, the content is used as the object, and the creator creating the content is used as the object producer, which is an example for illustration.

For one target content authored by a target author in this example, the behavior feature values of multiple dimensions of the target content include a content interaction feature value Score_com(the above object interaction feature value Score_com) And a browsing completion index value Score_fIndex value Score of browsing duration_durationIndex value Score of browsing times_vvAnd a browsing rate index value Score_str(ii) a The behavior characteristic values of the plurality of dimensions of the target content are obtained from history data of the content sharing platform, the history data includes behavior data generated by a history user who pushes the target content for the target content, and the behavior data may be collected by a terminal used by the history user and sent to a server of the content sharing platform.

In this example, a time period (hereinafter, referred to as the latest month) made up of a time from the current time to one month before the current time is described as an example of the current target time period; referring to fig. 3 and fig. 4, the following process is specifically included:

in step S301, the target content authored by the target author in the last month is determined.

Step S302, aiming at each determined target content, determining a content interaction characteristic value Score of each target content_comAnd a browsing completion index value Score_fIndex value Score of browsing duration_durationIndex value Score of browsing times_vvAnd a browsing rate index value Score_str。

That is, when the content in this example is a video, the content interaction feature value Score of each video created by the target creator may be determined according to the approval rate, the comment rate, and the sharing rate of each video created by the target creator_com(ii) a Determining browsing completion index value Score of each video according to the playing completion rate and the effective playing rate of the video created by each target creator_f(ii) a Determining a browsing duration index value Score of each video according to the playing duration of each video_duration(ii) a Determining a browsing frequency index value Score of each video according to the browsed frequency (i.e. browsing frequency) of each video_vv(ii) a Determining a browsing rate index value Score of each video according to the browsing rate of each video_str(ii) a The playing completion rate is the ratio of the actual watched time length of the video to the physical time length of the video; the effective playing rate is the ratio of the number of times that the playing completion rate of the video is greater than the playing completion rate threshold value to the total number of playing times. Further, the resource replacement characteristic value of each video can be determined according to the income of the electronic resource of each video, and further, the resource replacement characteristic value can be used for evaluating each video.

Step S303, determining a content interaction characteristic value Score_comAnd a browsing completion index value Score_fIndex value Score of browsing duration_durationIndex value Score of browsing times_vvAnd a browsing rate index value Score_strCorresponding weight w_com、w_f、w_d、w_vvAnd w_str。

In this example, the AHP method is used to determine the content interaction feature value Score_comAnd a browsing completion index value Score_fIndex value Score of browsing duration_durationIndex value Score of browsing times_vvAnd a browsing rate index value Score_strCorresponding weight w_com、w_f、w_d、w_vvAnd w_str。

It should be noted that step S303 may be a pre-processing procedure, and the content interaction feature value Score of the pre-processing procedure is specific to different target creators_comAnd a browsing completion index value Score_fIndex value Score of browsing duration_durationIndex value Score of browsing times_vvAnd a browsing rate index value Score_strCorresponding weight w_com、w_f、w_d、w_vvAnd w_strMay be identical.

Step S304, aiming at each target content, utilizing weight w_com、w_f、w_d、w_vvAnd w_strFor the content interaction feature value Score_comAnd a browsing completion index value Score_fIndex value Score of browsing duration_durationIndex value Score of browsing times_vvAnd a browsing rate index value Score_strAnd performing weighted summation to obtain content evaluation values (namely the object evaluation values) of the target contents.

In this step, a content evaluation value of each target content may be determined for each target content based on the following formula 16.

Score_i＝w_com×Score_{com_i}+w_f×Score_{f_i}+w_d×Score_{duration_i}+w_vv×Score_{vv_i}+w_str×Score_{str_i}Formula (16)

In formula 16, i is the identifier of the target content; score_iIs a content evaluation value of the target content identified as i; score_{com_i}Is a content interaction feature value of the target content identified as i; score_{f_i}Is a browsing completion index value of the target content identified as i; score_{duration_i}Is a browsing duration index value of the target content identified as i; score_{vv_i}Is a browsing number index value of the target content identified as i; score_{str_i}Is the browsing rate index value of the target content identified as i.

In step S305, the preset fifth evaluation reference value is used as a norm, and the content evaluation value of each target content is subjected to norm processing to obtain a content comprehensive evaluation value of the target creator (i.e., the target comprehensive evaluation value).

In the present example, the fifth evaluation reference value is set to 10, and the content evaluation values of the respective target contents are subjected to L (10) norm processing by the following formula 17.

In formula 17, Score is a comprehensive evaluation value of the content of the target creator; 10 is the above-mentioned fifth evaluation reference value; i is the identification of the target content authored by the target author; score_iFor the content evaluation value of the target content identified as i, N is the total amount of target content authored by the target author.

Step S306, obtaining the creation stability parameter of the target creator by using the preset confidence parameter and the total amount of the target content created by the target creator in the last month.

Reference is made specifically to the description related to equation 14 above, and the description is not repeated here.

Step S307, adjusting the comprehensive evaluation value of the content of the target creator by using the creation stability parameter to obtain the evaluation value of the target creator.

For details, reference is made to the description related to the above equation 15, and the description is not repeated here.

Referring to fig. 4, after the content sharing platform is evaluated by the creators in steps S301 and S307, the content sharing platform may be analyzed according to evaluation scores of the creators as follows:

1) author rating analysis

Determining the creator grade (namely the object grade) corresponding to each creator according to the evaluation score of each creator; from the user's perspective, the author level may reflect the degree of popularity of the content authored by the author from an a posteriori perspective (i.e., perspective of user feedback on the content); from the perspective of the content sharing platform, the level of the author can reflect the incentive direction of the platform target and the operation strategy of the content sharing platform to the author.

2) Content ecology evaluation

The content evaluation values of all the contents in the target time period on the content sharing platform can be aggregated to obtain the average value of the content evaluation values of all the contents; the evaluation scores of all creators in a target time period on the content sharing platform can be aggregated to obtain the mean value of the evaluation scores of all creators, and then the mean value of the content evaluation values of all contents and the mean value of the evaluation scores of all creators can be used as the evaluation index of the content analysis platform to evaluate the content ecology of the content analysis platform.

3) Authoring group analytics

The evaluation scores of a plurality of creators belonging to one creation group (MCN/guild) can be aggregated upwards to further obtain the evaluation index value of each creation group on the content sharing platform, and the creation capability or the popularity of each creation group is estimated according to the evaluation index value of each creation group.

4) Potential author analysis

The creator who has a chance to become a superior creator is determined as a potential creator (i.e., the potential object producer described above), for example, a creator having a larger increment of evaluation scores over a period of time than a threshold value is determined as a potential creator, or a creator having a larger increment of mean value of evaluation scores of content objects of the target content over a period of time is determined as a potential creator, etc.

In the example, the content evaluation values of a plurality of target contents created by a target creator are subjected to target fusion processing to obtain the evaluation result of the target creator, and the influence of the content evaluation values on the evaluation result of the creator, which meets the requirements of the content of a business target, is promoted in the target fusion processing, namely the influence of high-quality contents welcomed by a historical user on the evaluation result of the creator is promoted, so that the evaluation result of the creator, which meets the business target by the created contents, is obviously different from other creations, and the accuracy of evaluating the creator is obviously improved; in the example, based on the evaluation results of each author on the content sharing platform, author level analysis, content ecological analysis, authoring group analysis, potential author analysis and the like can be performed on the content sharing platform, and the business target of the content sharing platform operation can be further adjusted according to the analysis results, so that the popularity of the content sharing platform is improved.

In the embodiment of the application, the object evaluation value of the target object created by the target object producer is subjected to target fusion processing to obtain the evaluation result of the target object producer, and the influence of the target object of which the object evaluation value meets the business target on the evaluation result is promoted in the target fusion processing, so that the evaluation accuracy of the target object producer is promoted.

Referring to fig. 5, based on the same inventive concept, an embodiment of the present application provides an evaluation apparatus 500, including:

a first evaluating unit 501, configured to obtain object evaluation values of at least two target objects associated with a target object producer to be evaluated, where the object evaluation value of each target object is determined according to behavior feature values of multiple dimensions of a historical user, and the behavior feature values of the multiple dimensions are determined according to behavior data generated by the historical user for the target object;

the second evaluation unit 502 is configured to perform target fusion processing on the object evaluation values of the at least two target objects to obtain an evaluation result of the target object producer, where the target fusion processing includes an operation of improving an influence degree of the target object whose object evaluation value meets a business target on the evaluation result.

As an embodiment, the second evaluating unit 502 is specifically configured to:

weighting the object evaluation value of each of the at least two target objects by using a preset evaluation reference value to obtain a weighted object evaluation value of each target object, so that the first evaluation value deviation degree of the at least two target objects is higher than the second evaluation value deviation degree; the first evaluation value deviation degree is determined according to a deviation degree between the weighted object evaluation values of the first part of the object objects and the weighted object evaluation values of the second part of the object objects, and the second evaluation value deviation degree is determined according to a deviation degree between the object evaluation values of the first part of the object objects and the object evaluation values of the second part of the object objects; the first part of target objects comprise target objects of which the object evaluation values meet the business target, and the second part of target objects comprise target objects of the at least two target objects except the first part of target objects;

and fusing the weighted object evaluation reference values of the target objects to obtain the evaluation result of the target object producer.

As an embodiment, the second evaluating unit 502 is specifically configured to:

determining the evaluation reference value as a power, determining the power of the object evaluation value of each target object as a weighted object evaluation value of each target object, wherein the evaluation reference value is greater than 1;

the fusing the weighted object evaluation reference values of the target objects to obtain the evaluation result of the target object producer includes:

determining the evaluation reference value as a root index, and performing evolution processing on the sum of the weighted object evaluation values of all the target objects to obtain an object comprehensive evaluation value;

and obtaining the evaluation score of the target object producer based on the object comprehensive evaluation value.

As an embodiment, the second evaluating unit 502 is specifically configured to:

obtaining an authoring stability parameter of the target object producer by using a preset confidence coefficient parameter and the total number of the target objects based on the following formula, wherein the authoring stability parameter represents the stability of the target object producer in creating the number of the target objects meeting the business target:

in the above formula, R is an authoring stability parameter of the target object producer; e is a natural base number; n is the total number of the target objects; the CL is the confidence coefficient parameter; q is a preset confidence coefficient reference value;

and adjusting the object comprehensive evaluation value by using the creation stability parameter to obtain an evaluation score of the evaluation target object producer.

As an embodiment, the target objects whose object evaluation values meet the business target include at least two target objects, and after the target objects are sorted in the order from high to low according to the object evaluation values, the target objects in the top of the order are sorted; or

The target object whose object evaluation value meets the business target comprises the target object whose object evaluation value meets the object evaluation threshold value in the at least two target objects.

As an embodiment, the first evaluating unit 501 is specifically configured to:

for any one of the at least two target objects:

determining behavior characteristic values of multiple dimensions of the target object according to behavior data, generated by a historical user corresponding to the target object, aiming at the target object;

determining the importance of the behavior characteristic value of each dimension in the behavior characteristic values of the multiple dimensions based on the influence value between the behavior characteristic values of every two dimensions in the behavior characteristic values of the multiple dimensions;

and according to the importance of the behavior characteristic value of each dimension, carrying out weighting processing on the importance of the behavior characteristic value of each dimension to obtain an object evaluation value of the target object.

As an embodiment, the behavior feature values of the multiple dimensions include at least two feature values as follows:

an object interaction feature value; the object interaction characteristic value is determined according to the historical user and the data of the interactive operation of the target object;

an object browsing characteristic value; the object browsing characteristic value is determined according to the data of the target object browsed by the historical user;

a resource permutation eigenvalue; the resource replacement feature value is determined for the data of the target object transfer electronic resource based on the history user.

As an embodiment, the object browsing characteristic value includes at least one of the following index values:

a browsing completion index value obtained by processing a browsing completion rate of the target object, the browsing completion rate being determined based on a ratio of the browsing completion number to a total browsing number of the target object;

a browsing duration index value determined according to the duration information of the target object browsed by the historical user;

a browsing number index value determined by a total number of times the target object is browsed by the history user;

the browsing rate index value is determined based on the browsing rate of the target object, and the browsing rate is the ratio of the number of the historical users browsing the target object to the total number of the historical users corresponding to the target object.

As an example, the apparatus in fig. 5 may be used to implement any of the evaluation methods discussed above.

The evaluation apparatus 500 is a computer device shown in fig. 6 as an example of a hardware entity, and the computer device includes a processor 601, a storage medium 602, and at least one external communication interface 603; the processor 601, the storage medium 602, and the external communication interface 603 are connected by a bus 604.

The storage medium 602 has stored therein a computer program;

the processor 601, when executing the computer program, implements the method for testing the question-answering system of the test server 220 discussed above.

Fig. 6 illustrates an example of one processor 601, but the number of processors 601 is not limited in practice.

The storage medium 602 may be a volatile storage medium (volatile memory), such as a random-access memory (RAM); the storage medium 602 may also be a non-volatile storage medium (non-volatile memory), such as a read-only memory (rom), a flash memory (flash memory), a hard disk (HDD) or a solid-state drive (SSD), or the storage medium 602 may be any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited to this. The storage medium 602 may be a combination of the storage media described above.

Based on the same technical concept, the embodiment of the present application further provides a computer program product or a computer program, where the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, so that the computer device executes an evaluation method provided by the embodiment of the application.

Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

Alternatively, the integrated unit of the invention may be stored in a computer-readable storage medium if it is implemented in the form of a software functional module and sold or used as a separate product. Based on such understanding, the technical solutions of the embodiments of the present invention may be essentially implemented or a part contributing to the prior art may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the above methods of the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, a ROM, a RAM, a magnetic or optical disk, or various other media that can store program code.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims (10)

1. An evaluation method, comprising:

obtaining object evaluation values of at least two target objects related to a target object producer to be evaluated, wherein the object evaluation value of each target object is determined according to behavior characteristic values of multiple dimensions of a historical user, and the behavior characteristic values of the multiple dimensions are determined according to behavior data of the historical user aiming at the target object;

and performing target fusion processing on the object evaluation values of the at least two target objects to obtain an evaluation result of the target object producer, wherein the target fusion processing comprises an operation of improving the influence degree of the target object with the object evaluation value meeting the business target on the evaluation result.

2. The method according to claim 1, wherein performing target fusion processing on the object evaluation values of the at least two target objects to obtain the evaluation result of the target object producer comprises:

weighting the object evaluation value of each of the at least two target objects by using a preset evaluation reference value to obtain a weighted object evaluation value of each target object, so that the first evaluation value deviation degree of the at least two target objects is higher than the second evaluation value deviation degree; the first evaluation value deviation degree is determined according to the deviation degree of the weighted object evaluation values of the first part of object objects and the weighted object evaluation values of the second part of object objects, and the second evaluation value deviation degree is determined according to the deviation degree of the object evaluation values of the first part of object objects and the object evaluation values of the second part of object objects; the first part of target objects comprise target objects of which the object evaluation values meet the business target in the at least two target objects, and the second part of target objects comprise target objects except the first part of target objects in the at least two target objects;

and fusing the weighted object evaluation reference values of the target objects to obtain the evaluation result of the target object producer.

3. The method according to claim 2, wherein the weighting the object evaluation value of each of the at least two object objects by using a preset evaluation reference value to obtain a weighted object evaluation value of each object comprises:

determining the evaluation reference value as a power, determining the power of the object evaluation value of each target object as a weighted object evaluation value of each target object, wherein the evaluation reference value is greater than 1;

the fusing the weighted object evaluation reference values of the target objects to obtain the evaluation result of the target object producer comprises the following steps:

determining the evaluation reference value as a root index, and performing evolution processing on the sum of the weighted object evaluation values of all the target objects to obtain an object comprehensive evaluation value;

and obtaining the evaluation score of the target object producer based on the object comprehensive evaluation value.

4. The method of claim 3, wherein said deriving an evaluation score for said target subject producer based on said subject composite evaluation comprises:

obtaining an authoring stability parameter of the target object producer by using a preset confidence coefficient parameter and the total number of the target objects based on the following formula, wherein the authoring stability parameter represents the stability of the target object producer in authoring the number of the target objects meeting the business objective:

in the formula, the R is an authoring stability parameter of the target object producer; e is a natural base number; the N is the total number of the target objects; the CL is the confidence parameter; the q is a preset confidence coefficient reference value;

and adjusting the comprehensive evaluation value of the object by using the creation stability parameter to obtain the evaluation score of the evaluation target object producer.

5. The method according to any one of claims 1 to 4, wherein the target objects with the object evaluation values meeting the business objective comprise the target objects in the at least two target objects, which are ranked first after ranking the object evaluation values from high to low; or

The target object with the object evaluation value meeting the business target comprises at least two target objects, and the object evaluation value meets the target object of the object evaluation threshold.

6. The method according to any of claims 1-4, wherein said obtaining object evaluation values for at least two target objects of a target object producer to be evaluated comprises:

for any one of the at least two target objects:

determining behavior characteristic values of multiple dimensions of the target object according to behavior data, generated by a historical user corresponding to the target object, for the target object;

determining the importance of the behavior characteristic value of each dimension in the behavior characteristic values of the plurality of dimensions based on the influence value between the behavior characteristic values of every two dimensions in the behavior characteristic values of the plurality of dimensions;

and according to the importance of the behavior characteristic value of each dimension, carrying out weighting processing on the importance of the behavior characteristic value of each dimension to obtain an object evaluation value of the target object.

7. The method of any one of claims 1-4, wherein the behavior feature values for the plurality of dimensions include at least two feature values as follows:

an object interaction feature value; the object interaction characteristic value determines data of interactive operation on the target object according to the historical user;

an object browsing characteristic value; the object browsing characteristic value is determined according to the data of the target object browsed by the historical user;

a resource permutation eigenvalue; and determining the resource replacement characteristic value according to the historical user and aiming at the data of the target object transferred electronic resource.

8. An evaluation device, comprising:

the evaluation system comprises a first evaluation unit, a second evaluation unit and a third evaluation unit, wherein the first evaluation unit is used for obtaining object evaluation values of at least two object objects related to a target object producer to be evaluated, the object evaluation value of each object is determined according to behavior characteristic values of multiple dimensions of a historical user, and the behavior characteristic values of the multiple dimensions are determined according to behavior data generated by the historical user aiming at the object;

and the second evaluation unit is used for performing target fusion processing on the object evaluation values of the at least two target objects to obtain an evaluation result of the target object producer, and the target fusion processing comprises operation of improving the influence degree of the target object of which the object evaluation value meets the business target on the evaluation result.

9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1-7 are implemented when the program is executed by the processor.

10. A computer-readable storage medium having stored thereon computer instructions which, when executed on a computer, cause the computer to perform the method of any one of claims 1-7.

Images