Social consciousness user online incentive method and device based on two-stage game

Abstract

The invention provides a two-stage game-based social consciousness user online incentive method and equipment. The method comprises the following steps of: determining corresponding data centroid distribution for a plurality of task groups according to uploaded perception data; after determining a social network adjacency matrix of the mobile users in the task group, modeling a payment problem as a two-stage game of both parties, and determining utility functions of the mobile users and utility functions of service providers; and analyzing the optimal excitation of the service provider and the mobile user, and outputting an optimal strategy group with maximized effectiveness of the two parties. According to the method, the mutual combination of data quality detection, user personal utility, user social utility and user historical reputation is introduced to maintain the perception data quality and improve the utility of both parties at the same time.

Description

Technical Field

[0001] The present invention relates to the field of group intelligence perception, and in particular to a technical solution for designing an incentive mechanism based on a two-stage Stackelberg game. Background Art

[0002] With the development of sensor technology and network technology, mobile crowd sensing (MCS) technology plays an important role in task allocation and data collection in smart cities. A typical MCS system consists of a service provider and a group of mobile users. The service provider publishes various location-based sensing tasks and outsources them to users. One of the key challenges of crowd sensing is to design an effective incentive mechanism to ensure the quality and quantity of the collected data.

[0003] In recent years, the design of incentive mechanisms has been widely studied. Zhang et al. [1] proposed an online quality-aware incentive mechanism that effectively ensures the retention of quality incentives in the recruitment process. However, no abnormal data is processed. Hu et al. [2] proposed a combined zero-determinant (ZD) strategy that allows requesters to adopt a single method within their feasible range to reduce the overall prediction worker utility. Shi et al. [3] emphasized that social relationships are a key factor in designing agent incentives for crowdsourcing competitions. Zhang et al. [4] developed a multi-task allocation method that utilizes liquidity prediction. However, these strategies do not take social networks into account. The importance of social network effects is often significant because they have the potential to improve the effectiveness of MCS incentive mechanisms. Therefore, how to design an incentive mechanism and improve the quality of data and the utility of both parties while considering the social relationships and historical reputation of users is extremely difficult, so this field urgently needs to propose an online incentive technology solution that can improve data quality and the utility of both parties.

[0004] References:

[0005] [1] M.Zhang,

[0006] [2] Q.Hu, S.Wang, X.Cheng, J.Zhang, and W.Lv, "Cost-efficient mobilecrowdsensing with spatial-temporal awareness," IEEE Transactions on MobileComputing, vol.20, no.3, pp.928-938, 2021.

[0007] [3] Q.Shi and D.Hao, "Social sourcing: Incorporating social networks into crowdsourcing contest design," IEEE / ACM Transactions on Networking, vol.31, no.4, pp.1535-1549, 2023.

[0008] [4] J.Zhang and X.Zhang, "Multi-task allocation in mobile crowd sensing with mobility prediction," IEEE Transactions on Mobile Computing, vol.22, no.2, pp.1081-1094, 2023. Summary of the invention

[0009] In view of the above-mentioned defects in the prior art, the present invention provides a new mobile crowd intelligence perception incentive technology.

[0010] The present invention provides an online incentive method for social-conscious users based on a two-stage game, which determines the corresponding data centroid distribution for multiple task groups for uploaded perception data; after determining the social network adjacency matrix of mobile users in the task group, the payment problem is modeled as a two-stage game between the two parties, and the utility function of the mobile user and the utility function of the service provider are determined; the optimal incentives of the service provider and the mobile user are analyzed, and the optimal strategy group for maximizing the utility of both parties is output.

[0011] Moreover, the perception data uploaded by mobile users are preliminarily screened to remove the data that does not meet the threshold.

[0012] Moreover, the corresponding data centroid distribution is determined by performing Chebyshev distance calculation.

[0013] Moreover, the utility function of the mobile user is determined by calculating the data quality function, personal utility function, social utility function, user historical reputation function, reward function and cost.

[0014] Furthermore, the utility function of the service provider is determined by subtracting the reward function from the benefits brought by the total contributions of the users.

[0015] Furthermore, the optimal incentives of service providers and mobile users are analyzed using backward induction.

[0016] Moreover, the service provider collects the data and compiles it into data distribution. It measures the quality of the data uploaded by users by observing the deviation of data distribution, and updates the user reputation and rewards for the users.

[0017] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the social-aware user online incentive method based on the two-stage game as described above is implemented.

[0018] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned social-aware user online incentive method based on two-stage game.

[0019] On the other hand, the present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements the social awareness user online motivation method based on the two-stage game as described above.

[0020] In this invention, a mobile crowd intelligence perception incentive technology with socially aware users based on a two-stage Stackelberg game is proposed, the purpose of which is to improve the utility of mobile users and service providers by introducing users' social relationships, historical reputation, etc. under the premise of ensuring the quality of perception data. Compared with the prior art, the present invention has the following characteristics:

[0021] 1) Data screening function: By using the Chebyshev function to calculate the data distribution of the data uploaded by each mobile user, not only can false or erroneous data be screened out, but also the data quality utility of each user can be accurately calculated.

[0022] 2) Social relationship matrix: By constructing an adjacency matrix for all users participating in the task, social utility is provided to users who complete the same set of tasks, thereby increasing their own task rewards.

[0023] 3) Historical reputation function: According to the number of tasks completed by the user and the size of the perception data submitted, the average size of the perception data submitted by the user is estimated.

[0024] 4) Feasibility of the mechanism. Considering the selfish behaviors of both parties and the social benefits to users, the closed-form optimal strategies of both parties are derived. Theoretical analysis confirms the existence of a unique Stackelberg equilibrium (SE), ensuring that no participant will deviate from their optimal strategy.

[0025] Therefore, the present invention can ensure that the service provider consistently receives high-quality data while taking into account the utility of the user's social relationship, so that both parties can determine the optimal and unique strategy. In the task execution process, the existence of malicious users of different proportions is also taken into account, making the adaptation scenario more extensive. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a diagram illustrating an MCS scenario for a socially aware user according to an embodiment of the present invention.

[0027] Figure 2 This is a model diagram of the mobile crowd sensing (MCS) system according to an embodiment of the present invention.

[0028] Figure 3 This is a task selection and optimal strategy diagram for each mobile user according to an embodiment of the present invention.

[0029] Figure 4 The model diagram is a service provider optimal strategy and a mobile user optimal strategy of the SIM mechanism according to an embodiment of the present invention.

[0030] Figure 5 This is an experimental comparison chart of the optimal value of α1 in the embodiment of the present invention.

[0031] Figure 6 This is an experimental comparison chart of the optimal value of α2 in the embodiment of the present invention.

[0032] Figure 7 Embodiment a of the present invention i Experimental comparison chart of the optimal value of .

[0033] Figure 8 Embodiment b of the present invention i Experimental comparison chart of the optimal value of .

[0034] Fig. 9 FIG. 4 is a graph showing the total revenue of the service provider and the total return of all users according to an embodiment of the present invention.

[0035] Fig.10 Utility diagram of the service provider's revenue and mobile user's reward under different malicious ratios in an embodiment of the present invention.

[0036] Fig.11 Schematic diagram of user utility differences of different mechanisms under different malicious ratios according to an embodiment of the present invention.

[0037] Fig.12 This is a diagram of user uploaded data units under different malicious ratios in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The technical solution is described in detail below in conjunction with the accompanying drawings and embodiments.

[0039] The embodiment of the present invention proposes a mobile crowd intelligence perception incentive method for socially aware users based on a two-stage Stackelberg game. Figure 4 First, the perception data uploaded by mobile users is preliminarily screened, and the data that does not meet the threshold is removed. Then, the Chebyshev distance calculation is performed on the uploaded perception data for T task groups to determine T different data centroid distributions. After determining the social network adjacency matrix of all mobile users, the payment problem is modeled as a two-stage game between mobile users and service providers. The data quality function, personal utility, social utility, user historical reputation function, reward function, and cost function are calculated to determine the utility function of mobile users. The benefits brought by the total contribution of users are calculated, and the reward function is used to determine the utility function of the service provider. The optimal incentives of service providers and mobile users are analyzed by reverse induction, and the optimal strategy group that maximizes the utility of both parties is derived and output. Through theoretical analysis, the unique existence of Stackelberg equilibrium is proved, thus forming a win-win situation.

[0040] It can be considered that the present invention has established a new incentive mechanism, called the SIM mechanism. The mechanism introduces a combination of data quality detection, user personal utility, user social utility, and user historical reputation to maintain the perceived data quality while improving the utility of both parties in the game. The mechanism has the following characteristics:

[0041] 1. The SIM mechanism takes into account the existence of malicious users and provides malicious user identification and processing functions to ensure the authenticity and reliability of perception data.

[0042] 2. The SIM mechanism uses the Chebyshev distance function to filter the data by centroid, reducing the rewards for data distribution far from the centroid and increasing the rewards for data distribution close to the centroid.

[0043] 3. Create a diagonal matrix for the user group doing the same task to represent the direct social relationships of each user. Users can use social relationships to speed up task completion time, improve the quality of task completion, and thus increase their rewards.

[0044] See also Figure 1, the present invention is aimed at the scenario where users are responsible for collecting traffic data. In this system, the service provider recruits users, and mobile users use smart devices to collect traffic data according to their locations, and upload and submit them to the service provider. In this process, users can use social networks (or social networks) to speed up the completion of tasks and obtain higher rewards. There are two situations between mobile users: social relationships or non-social relationships. Users with social relationships are socially aware users. From the quality of uploaded data, users can be divided into normal socially aware users (i.e. socially aware users) and malicious users. Users doing the same task are divided into a group. In the same group, there may be both mobile users with social relationships and malicious users. The size of data uploaded by users doing the same task is different. The quality of data uploaded by users in a group is different, and there may be false data uploaded, thus affecting the income obtained by the service provider.

[0045] See also Figure 2 The embodiment provides a method for motivating online social awareness users based on a two-stage game. The implementation process can be considered to be based on the SIM mechanism system model and includes the following steps:

[0046] Step 1: Provide tasks. Service providers publish different types of tasks;

[0047] According to the dynamically randomly generated social network adjacency matrix, the utility functions of mobile users and service providers are determined respectively.

[0048] In the embodiment, the utility function of the mobile user is set as follows:

[0049] U i,j (x i,j ,X -i,j r i,j )=q i,j (x i,j ,w j )+f(x i,j )+φ(x i,j ,X i,j )+h(L i )+R i,j (x i,j )-cx i,j

[0050] Among them, U i，j (x i，j , X -i，j , r i，j ) represents the utility function of user i, X -i，j Indicates doing the same task t j In the group except x i，j The size of data submitted by other users other than q i，j (xi，j , w j ) represents the data quality function of user i, f(x i，j ) represents the personal utility function of user i, φ(x i，j , X i，j ) represents the social utility function of user i, h(L i ) represents the historical reputation function of user i, R i，j (x i，j ) represents the reward function of user i, c is the reward function of user i completing task t j cost parameters.

[0051] By introducing each function q i，j (x i，j , w j )、f(x i，j ),φ(x i，j , X i，j )、h(L i ), R i，j (x i，j ), can be rewritten as

[0052]

[0053] The utility function of the service provider:

[0054]

[0055] Among them, g ik Represents the social strength between user i and user k, and takes the value of 1 or 0. i，j represents the data quality function q of user i i，j (x i，j , w j ). i | represents the number of tasks that user i has participated in historically, and the absolute value represents a quantitative value. i，j Indicates that user i completes task t j Rewards. |P -i，j | represents the number of users other than user i, and the absolute value represents a quantity value. a i >0,b i >0, both represent the quality coefficients contributed by mobile users, which are different for different users due to the heterogeneity of devices. j (X j ) represents the utility function of the service provider, |P j | means to do the same task t jThe number of all users in the group, μ represents a parameter of the user's exchange equivalent monetary value, α1 and α2 are the two coefficients of the concave function that evaluates the effectiveness of the MCS service. In this way, the utility function calculation formula of the service provider is the first term to calculate the benefits brought by the total contribution of the user, and the second term is the reward function.

[0056] Step 2, Strategy and strategies The first strategy involves the service provider acting as a leader, completing tasks for the user. j Set reward r i，j The second strategy is that the user acts as a follower and receives a given unit reward r i，j , independently determine its own optimal submission data size x i，j , and take into account one's own social relationships.

[0057] The embodiment further provides an implementation method for determining the optimal strategy group. The two-stage game is analyzed using the reverse induction method, in which the service provider is the leader and the mobile user is the follower.

[0058] In the embodiment, first, the optimal upload perception data set of the mobile user is determined by analyzing the second stage

[0059]

[0060] Among them, U i，j represents the utility function of mobile users, Represents the first-order derivative of the mobile user's utility. If the first-order derivative is set to 0, the optimal upload perception data set can be obtained. for

[0061]

[0062] The proof process is as follows

[0063]

[0064] Among them, the optimal uploaded perception data set Indicates that user i completes task t j The optimal strategy set of Indicates that user k completes task t j The optimal strategy set. i w j +b i +r i，j -c| represents taking the absolute value of the four operations of these numbers, and there are

[0065]

[0066] in, represents the second-order derivative of the mobile user utility function, represents the second-order partial derivative of the mobile user utility function. represents the optimal unit reward set of the service provider. Similarly, the optimal unit reward of the service provider for

[0067]

[0068] The proof process is as follows

[0069] By bringing in The utility function of the service provider and its first-order derivative can be obtained

[0070]

[0071] represents the first-order derivative of the service provider utility function, and the variable η represents make is 0, the optimal unit reward set can be solved

[0072] Because there is

[0073]

[0074] Representing the second-order derivative of the service provider's utility function, we can get Existing and unique.

[0075] Finally, it was proved For the optimal strategy group, a unique SE is formed, which also indicates that both parties have no motivation to unilaterally deviate from their optimal strategies.

[0076] Step 3, completing the perceived task. The service provider informs the user of the assigned task and provides information about the compensation for completing the task. Users evaluate the costs and potential risks associated with the task taking into account their social relationships. Based on this evaluation, they decide whether to accept the task and upload data in exchange for a reward.

[0077] Step 4: Analyze the perception data. The service provider collects the data and compiles it into data distribution. The quality of the data uploaded by the user is measured by observing the data distribution deviation.

[0078] Furthermore, the embodiment of the present invention proposes a specific analysis implementation method as follows:

[0079] 4.1 Preliminary screening of perception data uploaded by mobile users.

[0080] The embodiment uses d(x i，j , δj ) function for preliminary screening, where d(x i，j , δ j ) represents the data filtering function, x i，j Indicates that mobile user i completes sensing task t j The size of the uploaded perception data, δ j Indicates the threshold of the perceived task size. The value can be set according to the actual situation during implementation. In the embodiment, it is set to δ j =0, as shown below.

[0081]

[0082] 4.2 The Chebyshev distance is used to calculate the distribution of data centroids for different task groups for the uploaded perception data.

[0083] After preliminary screening, the uploaded perception data is used to calculate the distribution of data centroids for different task groups using Chebyshev distance. Let w(X j ) represents the centroid function of data distribution, X j represents a group of mobile users who complete the same task, D Chebyshev (x i，j , x k，j ) indicates that the calculation is to do the same task t j The Chebyshev distance function between user i and user k, x k，j Indicates that user k completes task t j The size of the uploaded perceived data, z is used to identify x i，j The dimension of (the embodiment sets z = 1), Indicates that user i completes task t in dimension z j , Similarly, it means that user k completes task t in dimension z j ∈ represents a belonging relationship. The data distribution centroid function is expressed as follows.

[0084]

[0085] Among them, w j Represents task t j The data centroid value, w(X j ) aims to find the variable w that minimizes the Chebyshev distance function j The value of .

[0086] By analyzing the size of data quality or comparing the data quality centroid, false or erroneous data can be screened out, which is helpful for identifying malicious users.

[0087] Step 5: Update user reputation. The reputation of user i is determined by the service provider based on the completion of the task, and the reward brought by reputation increases with the size of the perceived data of each task.

[0088] Step 6: Update rewards for users. To motivate the generation of rewards, the service provider adopts a two-stage Stackelberg game approach, which aims to determine individual rewards and the total income of all users to achieve Nash equilibrium (NE). By considering indicators such as data quality, private utility, social utility, historical reputation, etc., the service provider can assign the best reward to any user in the MCS system to continue to promote the completion of the task.

[0089] See also Figure 3 In the example provided, there are 3 mobile users (represented by p), 2 tasks (represented by t), g represents the influence of social relationships, and the optimal reward strategy for mobile users to complete tasks and optimal upload data size strategy The specific information provided is shown in Table 1, which shows 3 mobile users, 2 tasks, and their social networks. and The result calculated by the formula.

[0090] Table 1

[0091]

[0092] In order to facilitate the explanation of the technical effect of the present invention, a comparative experiment is conducted using the technical solution of the present invention and the prior art as follows:

[0093] 1) Determine the values ​​of each parameter. Figure 5 , 6, 7, and 8, provide comparisons under three different parameters. The horizontal axis is the number of tasks, ranging from 1 to 10.

[0094] For a i For example, i The smaller the value of , the larger the corresponding overall return and reward value, and the larger the gap between the two. In the formula, a i As the denominator. i The larger the value, the stricter the constraints, and the lower the total benefits and rewards. On the contrary, loose constraints can effectively stimulate users to participate in tasks and upload data. i with a i On the contrary, when b i When the value of is larger, the total benefit and reward increase. It can be seen that it is preferred to choose a smaller a i and larger b i To maintain mission teamwork.

[0095] For α1 and α2, when the value of α1 is large, the total income and total reward increase significantly, and the difference between the two also increases significantly. Therefore, choosing a larger α1 value can increase the possibility of users uploading data. The larger α2 is, the diminishing marginal benefit reduces the total income and reward, so choosing a smaller α2 is beneficial to the user's income.

[0096] 2) Compare the revenue and rewards of each mechanism without malicious users.

[0097] As attached Fig. 9 As shown, the performance of the five mechanisms is compared from the total revenue of the service provider and the total reward among users. Among them, QnQ, POM, CPSS and Social are existing mechanisms, and SIM represents the mechanism provided by the present invention. QnQ, POM and SIM all perform well, but the total revenue and total user reward of CPSS are the lowest. Among them, the total revenue and total reward of SIM are slightly higher than those of QnQ because both adopt a two-stage Stackelberg game. In game-based teamwork, service providers try to reach a Nash equilibrium by evenly distributing tasks among users. Since the mechanism does not take social networks into account, the total reward and total revenue of POM are lower.

[0098] 3) Compare the revenue and rewards of each mechanism under different proportions of malicious users.

[0099] As attached Fig.10 As shown in Figure 2, the revenue, reward, and utility of each mechanism are compared when the ratio of total reward to malicious users is 0.3, 0.5, and 0.7. The total revenue and total reward of SIM, QnQ, and POM decrease as the ratio of malicious users increases. However, in comparison, SIM still has the highest total revenue and total reward. It is worth noting that the results of Social are abnormal, and the total revenue and total reward increase with the increase of malicious user rate.

[0100] Fig.11 The difference in user utility of different mechanisms under different malicious ratios is illustrated, where the horizontal axis represents the ratio of malicious users 0.3, 0.5, and 0.7. The results show that SIM has the highest performance, followed by QnQ, POM, and CPSS.

[0101] 4) Compare the uploaded data units of each mechanism under different proportions of malicious users.

[0102] As attached Fig.12As shown in the figure, the amount of data uploaded by malicious users changes from 0.3 to 0.7. The amount of uploaded data of SIM, QnQ and POM gradually decreases as the malicious ratio increases, while the amount of uploaded data of CPSS remains unchanged because it does not consider the quality and quantity of data. SIM outperforms other mechanisms because it considers data quality and user reputation.

[0103] During specific implementation, the above process can be automatically executed using computer software technology.

[0104] In addition, the present application also relates to an electronic device, comprising at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned social-aware user online incentive method based on two-stage game.

[0105] The present application also relates to a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to execute the above-mentioned social awareness user online incentive method based on two-stage game.

[0106] The present application also relates to a computer program product, including a computer program, characterized in that when the computer program is executed by a processor, the above-mentioned social awareness user online incentive method based on two-stage game is implemented.

[0107] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0108] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for motivating socially conscious users online based on a two-stage game, characterized in that: Determine the corresponding data centroid distribution for multiple task groups for the uploaded perception data; after determining the social network adjacency matrix of mobile users in the task group, model the payment problem as a two-stage game between the two parties, and determine the utility function of the mobile user and the utility function of the service provider; analyze the optimal incentives of the service provider and the mobile user, and output the optimal strategy group that maximizes the utility of both parties.

2. According to claim 1, a method for motivating online social awareness users based on a two-stage game, characterized in that: Perform a preliminary screening of the perception data uploaded by mobile users and remove the data that does not meet the threshold.

3. The method for online motivation of social awareness users based on two-stage game according to claim 1, characterized in that: The corresponding data centroid distribution is determined by performing Chebyshev distance calculation.

4. The method for online motivation of social awareness users based on two-stage game according to claim 1 is characterized by: The utility function of mobile users is determined by calculating data quality function, personal utility function, social utility function, user historical reputation function, reward function and cost.

5. The method for online motivation of social awareness users based on two-stage game according to claim 1 is characterized by: The utility function of the service provider is determined by subtracting the reward function from the total contribution of the user.

6. The method for online motivation of social awareness users based on two-stage game according to claim 1, characterized in that: The optimal incentives of service providers and mobile users are analyzed using backward induction.

7. The method for online motivation of social awareness users based on two-stage game according to claim 1, characterized in that: The service provider collects data and compiles statistics into data distribution. It measures the quality of data uploaded by users by observing data distribution deviation, and updates user reputation and rewards for users.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the online motivation method for social-aware users based on two-stage game as described in any one of claims 1 to 7 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for online motivation of socially conscious users based on two-stage game as described in any one of claims 1 to 7 is implemented.

10. A computer program product, comprising a computer program, characterized in that: When the computer program is executed by a processor, the method for online motivation of socially conscious users based on two-stage game as described in any one of claims 1 to 7 is implemented.

Images