Sorting Method, Device and Computer Equipment Based on Multi-Stage Viewpoint Interaction

Abstract

This application relates to a sorting method, device, and computer equipment based on multi-stage view interaction. The method includes: determining a quadruple of users in a social network and the interactive learning objects of the users. In the current interaction round, according to a preset learning rate and the sorting, calculate the scoring contribution of the items to be sorted for the users. According to the preset learning rate, the familiarity of the interactive learning objects, and the sorting of the interactive learning objects, calculate the aggregated scoring contribution of the set of interactive learning objects. Update the sorting according to the scoring contribution and the aggregated scoring contribution to obtain a learning sorting. Update the preference degree of the learning sorting according to the preference and the preference stubbornness to obtain an updated sorting. When the consistency ratio of the updated sorting of the users reaches a threshold or the maximum number of interaction iterations is reached, output a consistent sorting result. Using this method can perform multi-user view fusion and sorting result output more realistically and flexibly.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular, to a sorting method, device, and computer device based on multi-stage view interaction. Background Art

[0002] Group Decision Making (GDM) involves a collaborative process in which multiple individuals gather together to collect a wide range of ideas, integrate various opinions, and reach a unified decision. By integrating the opinions and knowledge of multiple individuals, GDM can achieve more accurate and robust decision-making results in complex environments. GDM methods generally include two steps: the Consensus Reaching Process (CRP) and the solution selection process. In the initial stage of group decision-making, there are often conflicts or significant differences in the opinions of users. Therefore, users need to adjust their views through an update and iteration process in the CRP to improve the consensus level of the group. Therefore, the CRP is necessary for achieving decision-making consistency among most users. In the CRP, users are not completely independent, especially due to the rapid development of instant messaging tools (such as WeChat, WhatsApp, Facebook) and social media platforms (such as Instagram, Weibo, Twitter). These platforms have significantly increased the frequency of social interaction and profoundly affected the social networks and relationship structures of individuals.

[0003] However, existing rules for the evolution process of group decision-making opinions mostly focus on the field of opinion dynamics and rarely consider knowledge in sociology and psychology. Although the designed opinion interaction rules can reach a consensus relatively quickly, it is difficult to reflect the actual object selection process and opinion interaction evolution rules in the CRP. Summary of the Invention

[0004] Based on this, it is necessary to provide a sorting method, device, and computer device based on multi-stage view interaction for the above technical problems.

[0005] A sorting method based on multi-stage view interaction, the method includes:

[0006] Determine the quadruple of users in the social network and the interactive learning objects of the users; wherein, the quadruple includes: familiarity with the item to be sorted, preference, preference stubbornness, and sorting; the interactive learning objects are determined according to the similarity of users in the social network;

[0007] In the current interaction round, according to the preset learning rate and the said ranking, calculate the scoring contribution of the items to be ranked for the user, and according to the preset learning rate, the familiarity of the interactive learning objects, and the ranking of the interactive learning objects, calculate the aggregated scoring contribution of the set of interactive learning objects;

[0008] Update the said ranking according to the said scoring contribution and the aggregated scoring contribution to obtain a learning ranking;

[0009] Update the preference degree of the learning ranking according to the said preference and the preference stubbornness to obtain an updated ranking;

[0010] When the consistency ratio of the updated rankings of all users reaches the threshold, or reaches the maximum interaction iteration number, output the consistency ranking result.

[0011] In one of the embodiments, it further includes: obtaining the ranking of the user and the ranking of the user to be evaluated , where M is the total number of items to be ranked;

[0012] Calculate the normalized probability distributions P and Q of the rankings and the ranking respectively as:

[0013] ;

[0014] ;

[0015] where, ;

[0016] Calculate the KL divergence of the probability distribution P with respect to the average probability distribution and the KL divergence of the probability distribution Q with respect to the average probability distribution respectively as:

[0017] ;

[0018] ;

[0019] where, ;

[0020] Calculate the JS divergence according to the KL divergence and the KL divergence as:

[0021] .

[0022] Convert the JS divergence to the JS distance:

[0023] 。

[0024] When the JS distance between the user and the user to be evaluated is less than the threshold, the user to be evaluated is determined as an interactive learning object.

[0025] In one embodiment, it further includes: calculating the scoring contribution of the item to be sorted for the user according to the preset learning rate and the sorting as:

[0026] ;

[0027] Wherein, represents the scoring contribution, represents the position of item i in the sorting, M is the total number of items in the sorting, represents the learning rate.

[0028] In one embodiment, it further includes: calculating the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning objects, and the sorting of the interactive learning objects as:

[0029] ;

[0030] Wherein, represents the familiarity weight of the interactive learning object , K is the total number of interactive learning objects, represents the familiarity of the interactive learning object , represents the ranking of item i for the interactive learning object .

[0031] In one embodiment, it further includes: adding the scoring contribution and the aggregated scoring contribution to obtain a comprehensive score as:

[0032] 。

[0033] According to the comprehensive score, the items in the sorting are sorted in descending order to obtain a learning sorting.

[0034] In one embodiment, the preferences include: absolute preference and relative preference; wherein, the absolute preference represents an absolute constraint on the sorting of items in the sorting, and the relative preference represents a constraint on the relative sorting between items in the sorting.

[0035] In one embodiment, it further includes: the preference stubbornness is initially 1, and when it is necessary to execute according to the preference stubbornness, the preference stubbornness decreases after execution.

[0036] A sorting device based on multi-stage view interaction, the device comprising:

[0037] An initialization module for determining a quadruple of a user in a social network and an interactive learning object of the user; wherein, the quadruple includes: familiarity with the item to be sorted, preference, preference stubbornness, and sorting; the interactive learning object is determined according to the similarity of users in the social network;

[0038] A scoring calculation module for calculating the scoring contribution of the item to be sorted of the user according to a preset learning rate and the sorting in the current interaction round, and calculating the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning object, and the sorting of the interactive learning object;

[0039] A learning update module for updating the sorting according to the scoring contribution and the aggregated scoring contribution to obtain a learning sorting;

[0040] A preference setting module for updating the preference degree of the learning sorting according to the preference and the preference stubbornness to obtain an updated sorting;

[0041] A sorting output module for outputting a consistent sorting result when the consistency ratio of the updated sortings of all users reaches a threshold or the maximum number of interaction iterations is reached.

[0042] A computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program:

[0043] Determining a quadruple of a user in a social network and an interactive learning object of the user; wherein, the quadruple includes: familiarity with the item to be sorted, preference, preference stubbornness, and sorting; the interactive learning object is determined according to the similarity of users in the social network;

[0044] In the current interaction round, calculating the scoring contribution of the item to be sorted of the user according to a preset learning rate and the sorting, and calculating the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning object, and the sorting of the interactive learning object;

[0045] Updating the sorting according to the scoring contribution and the aggregated scoring contribution to obtain a learning sorting;

[0046] Updating the preference degree of the learning sorting according to the preference and the preference stubbornness to obtain an updated sorting;

[0047] When the consistency ratio of the updated sortings of all users reaches a threshold or the maximum number of interaction iterations is reached, outputting a consistent sorting result.

[0048] The above sorting method, device, and computer device based on multi-stage view interaction first determine the quadruple of users in the social network and the interactive learning objects of the users. The familiarity is introduced into the quadruple, which reflects the choices and learning processes of users under different familiarity levels in the real consensus reaching. At the same time, the preference and preference stubbornness are introduced, so that the preference of the user can be executed during view fusion, thereby ensuring the rationality and consistency of the sorting result. Based on the above initialization design, during interaction, first calculate the score of the current sorting and the score of the interactive learning object sorting. When calculating the score, the concept of familiarity is introduced, so that the opinions of users with higher familiarity occupy a larger proportion in the final view update. Finally, update the sorting according to the preference and preference stubbornness, and achieve a consistent sorting result through multiple rounds of iteration. Therefore, the above solution of the present application can perform multi-user view fusion and sorting result output more realistically and flexibly. Brief Description of the Drawings

[0049] Figure 1 It is a schematic flowchart of a sorting method based on multi-stage view interaction in an embodiment;

[0050] Figure 2 It is a structural block diagram of a sorting device based on multi-stage view interaction in an embodiment;

[0051] Figure 3 It is an internal structure diagram of a computer device in an embodiment. Detailed Description of the Embodiment

[0052] In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0053] In one embodiment, as Figure 1 shown, a sorting method based on multi-stage view interaction is provided, including the following steps:

[0054] Step 102, determine the quadruple of users in the social network and the interactive learning objects of the users.

[0055] The social network can be modeled as an undirected graph where represents the set of edges, represents the set of nodes. The edges in the undirected graph represent the connections between users, and the nodes correspond to the users. The adjacency matrix of graph is where:

[0056] ;

[0057] Indicates the user and There is a connection between them, and they can interact during the decision-making process.

[0058] In an undirected graph, a user has four attributes, which are represented by a quadruple. The quadruple includes: familiarity with the items to be sorted, preference, preference stubbornness, and sorting. Specifically, users with higher familiarity are given higher weights during the interactive iteration process. Preference describes the user's tendency towards items, and preference stubbornness measures the degree to which the user adheres to their preference. Sorting represents the user's sorting of multiple items, which is updated after each iteration round.

[0059] The interactive learning object is determined based on the similarity of users in the social network.

[0060] Step 104, in the current interactive round, calculate the scoring contribution of the items to be sorted by the user according to the preset learning rate and sorting, and calculate the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning object, and the sorting of the interactive learning object.

[0061] Step 106, update the sorting according to the scoring contribution and the aggregated scoring contribution to obtain the learning sorting.

[0062] Step 108, update the preference degree of the learning sorting according to the preference and the preference stubbornness to obtain the updated sorting.

[0063] Step 110, when the consistency ratio of the updated sortings of all users reaches the threshold or the maximum number of interactive iterations is reached, output the consistent sorting result.

[0064] In the above sorting method based on multi-stage view interaction, first, determine the quadruple of users in the social network and the interactive learning objects of the users. Familiarity is introduced in the quadruple, which reflects the selection and learning process of users under different familiarity levels in the real consensus reaching. At the same time, preference and preference stubbornness are introduced, which can execute the user's preference during view fusion, thus ensuring the rationality and consistency of the sorting result. Based on the above initialization design, during the interaction, first calculate the score of the current sorting and the score of the interactive learning object sorting. When calculating the score, the concept of familiarity is introduced, so that the opinions of users with higher familiarity account for a larger proportion in the final view update. Finally, update the sorting according to the preference and the preference stubbornness, and reach the consistent sorting result through multiple rounds of iteration. Therefore, the above solution of the present application can perform multi-user view fusion and sorting result output more realistically and flexibly.

[0065] In one embodiment, the interactive learning object is determined according to the similarity of users in the social network, specifically including:

[0066] S1. Obtain the ranking of the user and the ranking of the user to be evaluated , where M is the total number of items to be ranked.

[0067] S2. Calculate the normalized probability distributions P and Q of the rankings and the ranking respectively as follows:

[0068] ;

[0069] ;

[0070] where .

[0071] S3. Calculate the KL divergence of the probability distribution P with respect to the average probability distribution and the KL divergence of the probability distribution Q with respect to the average probability distribution :

[0072] ;

[0073] ;

[0074] where ;

[0075] S4. Calculate the JS divergence according to the KL divergence and the KL divergence as follows:

[0076] .

[0077] S5. Convert the JS divergence to the JS distance:

[0078] .

[0079] S6. When the JS distance between the user and the user to be evaluated is less than the threshold, determine the user to be evaluated as an interactive learning object.

[0080] In this embodiment, since the JS divergence itself is a divergence metric in information theory, its value range is . To convert it into a more intuitive distance metric, after taking the square root, the value range of the JS distance is .

[0081] After the user initializes the quadruple, the user is placed in a preset social network. The user can interact with the neighbors who have established connections in the network. However, considering the interaction process in reality, when selecting the interaction object, the user will give priority to the consistency of the neighbor's opinion with his own opinion, and tend to communicate with users with smaller differences in opinions. Based on this, a method for selecting interactive neighbors based on label similarity is proposed. When selecting interactive neighbors, the user first evaluates the similarity distance with each neighbor according to the user ranking label similarity calculation method based on JS distance proposed in . After completing the distance calculation, the distance is compared with the preset threshold similar_threshold. Those neighbor users whose JS distance with themselves is less than the threshold similar_threshold are determined as interactive learning objects, and the selected users among these neighbors are stored in the user person_similar set.

[0082] After determining the interactive object set user person_similar, the user's familiarity label needs to be considered at this time. The familiarity label F is used to measure the user's familiarity with a specific item or field, and is a key factor affecting the weight of the neighbor's opinion being learned. The reason for introducing familiarity is that users have a deeper understanding and more reliable judgment in their familiar fields, so their opinions should contribute more to consensus. Familiarity F is usually evaluated and determined through data in multiple dimensions, including the user's historical performance, professional background, number of projects participated in, and frequency of interaction with specific items. For example, indicators such as the number of papers published by a user in a related field, the number of research projects participated in, and the degree of activity in the user's social network can all be used to quantify their familiarity. In addition, familiarity can also be supplemented by methods such as questionnaires or user self-evaluation to ensure that it can accurately reflect the user's actual familiarity. In this embodiment, the familiarity label F will be used to weight the neighbors' opinions so that the opinions of users with higher familiarity occupy a larger proportion in the final opinion update, thereby improving the accuracy of opinion fusion and the efficiency of consensus formation.

[0083] In addition, when users interact with each other, there is a certain learning ratio and retention part between their original opinions and the opinions obtained from their neighbors. Therefore, this study introduces the learning rate To reflect the degree to which the user adopts the neighbor's opinions during the opinion update process. Learning rate The weight distribution of users when fusing their own opinions with those of their neighbors is controlled, so that users can dynamically adjust their acceptance of new information. , can balance the relationship between users retaining their original opinions and accepting new opinions from their neighbors, thereby improving the flexibility and accuracy of opinion updating.

[0084] The ranking list of the user himself / herself is denoted as , where represents the item number of the user ranked at the k-th position. The set of similar neighbors (interactive learning objects) is denoted as , where K is the total number of similar neighbors, and each neighbor has a familiarity , and its ranking list is denoted as , where represents the item number of neighbor ranked at the k-th position. The learning rate is denoted as , which represents the proportion of the user's learning of the neighbor's opinion in the final opinion during the interaction process.

[0085] Specifically, the calculation method of the scoring contribution is as follows:

[0086] According to the pre-set learning rate and sorting, the scoring contribution of the item to be sorted by the user is:

[0087] ;

[0088] where represents the scoring contribution, represents the position of item i in the said sorting, M is the total number of items in the sorting.

[0089] The calculation method of the aggregated scoring contribution is:

[0090] According to the pre-set learning rate, the familiarity of the interactive learning object, and the sorting of the interactive learning object, the aggregated scoring contribution of the set of interactive learning objects is:

[0091] ;

[0092] where represents the familiarity weight of the interactive learning object , K is the total number of interactive learning objects, represents the familiarity of the interactive learning object , represents the ranking of item i for the interactive learning object .

[0093] When performing fusion, the scoring contribution and the aggregated scoring contribution are added together to obtain the comprehensive score as:

[0094] ;

[0095] The comprehensive score is specifically expanded as:

[0096] According to the comprehensive score, the items in the ranking are sorted in descending order to obtain the learning ranking, specifically:

[0097] ,

[0098] ;

[0099] Among them .

[0100] In one embodiment, the preferences include: absolute preference and relative preference; wherein, the absolute preference represents an absolute constraint on the ranking of items in the ranking, and the relative preference represents a constraint on the relative ranking between items in the ranking.

[0101] After the user combines their initial ranking with the neighbor learning to obtain a new ranking, it is also necessary to comprehensively consider the user's preference attributes. As mentioned before, it represents the user's preference for items, and the preference is divided into two forms: the first is the absolute preference, that is, the user prefers a certain item to be in the top C% of the ranking; the second is the relative preference, that is, the user believes that the ranking of item i must be before item j. Specifically, the absolute preference requires a certain item to be in the top C% position in the total ranking. For example, in the case of five items, if a user prefers item 1 to be in the top 60%, then the ranking of item 1 is limited to the 1st, 2nd, and 3rd positions. The relative preference stipulates the ranking order between items. For example, a user believes that the ranking of item A must be before item B.

[0102] In another embodiment, the initial value of the preference stubbornness B is set to 1, indicating that the user fully adheres to their preferences in the initial stage. During the iterative process, when the new ranking result does not meet the user's preference requirements, preference enforcement is required, and the preference stubbornness is adjusted accordingly. The preference stubbornness B, as a probability value, determines the way of preference enforcement. Initially, B = 1, that is, the preference enforcement is deterministic. Whenever preference enforcement is required, the user enforces it with a probability of the preference stubbornness B, and B is reduced according to a predetermined decay mechanism after each execution, thereby gradually changing the preference enforcement from deterministic to probabilistic.

[0103] For users with absolute preferences, if the new ranking obtained after learning does not meet the preference requirements, the preferred item needs to be randomly placed in a certain position in the top C% of the ranking. For users with relative preferences, if the new ranking does not meet the preference requirements, the positions of the two items with the preferred order need to be exchanged to ensure that item i is ranked before item j.

[0104] Through the above mechanism, not only can the user's view be dynamically adjusted to integrate the opinions of neighbors, but also the user's preferences can be enforced when necessary, thus ensuring the rationality and consistency of the sorting results. At the same time, as the preference stubbornness gradually decreases, the enforcement of the user's preferences will gradually change to probabilistic execution, enhancing the flexibility and adaptability of the model in the process of diverse opinion integration.

[0105] In the above manner, in each iteration, the system traverses each user in turn, performs the steps of view update and sorting adjustment, and then updates its sorting attributes. Specifically, each user first generates a new sorted list according to the aggregated score and preference correction, and then checks whether the new sorting meets its preference requirements. If not, preference enforcement is performed and the preference stubbornness is adjusted. The entire iterative process continues until either of the following termination conditions is met: the consistency ratio of the sorting labels of all users in the overall population reaches a preset threshold, or the predetermined maximum number of iterations is reached. Through multiple rounds of iteration, the opinions in the user social network gradually tend to be consistent, and finally a consensus conclusion is formed.

[0106] In summary, the present application can be applied in the following application scenarios. A certain megacity needs to determine the annual investment priorities of five types of new energy infrastructure, which are: distributed photovoltaic power grid, electric vehicle fast charging network, hydrogen production center, intelligent energy storage system, and biomass power generation station. The five types of new energy infrastructure focus on environmental protection orientation, people's livelihood needs, technological foresight, and grid stability respectively.

[0107] Nine experts use to represent, and the five types of new energy infrastructure use to represent. The familiarity of experts, preference types and specific preferences as well as the initial sorting are generated by a random generation method. The preference stubbornness is initialized to 1. The randomly initialized parameters in this example are shown in Table 1.

[0108] Table 1 Parameter Design and Results after Iteration

[0109]

[0110] Nine experts are distributed on a two-dimensional grid. Each expert needs to rank the five types of new energy infrastructure to reflect their individual preferences and the adjustment of views affected by neighbors. The preference stubbornness parameter of the expert has an initial value of 1.0. After each preference is enforced, the stubbornness decreases by a decay factor of 0.9. The initial ranking list of each expert Generated randomly and adjusted according to its preferences during the initialization phase to meet its inherent preferences. The consensus threshold is set to 0.9, and the maximum number of iterations is set to 20 times, and the similarity threshold is set to 0.8, and the learning rate is set to 0.9. Then, the embodiments of the present application are adopted, and finally, after multiple rounds of discussion, information integration, and feedback adjustment, the sorting result is output.

[0111] It should be understood that although Figure 1 the steps in the flowchart of Figure 1 are shown in sequence according to the indication of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover,

[0112] In one embodiment, as Figure 2 shown, a sorting device based on multi-stage view interaction is provided, including: an initialization module 202, a scoring calculation module 204, a learning update module 206, a preference setting module 208, and a sorting output module 210, where:

[0113] The initialization module 202 is used to determine the quadruple of the user in the social network and the user's interactive learning object; wherein, the quadruple includes: familiarity with the item to be sorted, preference, preference stubbornness, and sorting; the interactive learning object is determined according to the similarity of users in the social network;

[0114] The scoring calculation module 204 is used to calculate the scoring contribution of the item to be sorted by the user according to the preset learning rate and the sorting in the current interaction round, and calculate the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning object, and the sorting of the interactive learning object;

[0115] The learning update module 206 is used to update the sorting according to the scoring contribution and the aggregated scoring contribution to obtain a learning sorting;

[0116] The preference setting module 208 is used to update the preference degree of the learning sorting according to the preference and the preference stubbornness to obtain an updated sorting;

[0117] A sorting output module 210, configured to output a consistent sorting result when the consistency ratio of the updated sorting of all users reaches a threshold or the maximum number of interactive iterations is reached.

[0118] In one embodiment, the initialization module 202 is further configured to obtain the sorting of the user and the sorting of the user to be evaluated , where M is the total number of items to be sorted;

[0119] Calculate the normalized probability distributions P and Q of the sorting and the sorting respectively as:

[0120] ;

[0121] ;

[0122] Among them, ;

[0123] Calculate the KL divergence of the probability distribution P with respect to the average probability distribution and the KL divergence of the probability distribution Q with respect to the average probability distribution :

[0124] ;

[0125] ;

[0126] Among them, ;

[0127] According to the KL divergence and the KL divergence calculate the JS divergence as:

[0128] ;

[0129] Convert the JS divergence to the JS distance:

[0130] .

[0131] When the JS distance between the user and the user to be evaluated is less than the threshold, determine that the user to be evaluated is an interactive learning object.

[0132] In one embodiment, the scoring calculation module 204 is further configured to calculate the scoring contribution of the items to be sorted by the user according to a preset learning rate and the sorting as:

[0133] ;

[0134] Among them, represents the scoring contribution, represents the position of item i in the said sorting, M is the total number of items in the sorting, represents the learning rate.

[0135] In one of the embodiments, the scoring calculation module 204 is further configured to calculate the aggregated scoring contribution of the set of interactive learning objects according to a preset learning rate, the familiarity of the interactive learning objects, and the sorting of the interactive learning objects as:

[0136] ;

[0137] Among them, represents the familiarity weight of the interactive learning object , K is the total number of interactive learning objects, represents the familiarity of the interactive learning object , represents the ranking of item i for the interactive learning object .

[0138] In one of the embodiments, the scoring calculation module 204 is further configured to add the said scoring contribution and the aggregated scoring contribution to obtain the comprehensive score as:

[0139] .

[0140] According to the said comprehensive score, the items in the said sorting are sorted in descending order to obtain the learning sorting.

[0141] In one of the embodiments, the preferences include: absolute preference and relative preference; among them, the absolute preference represents the absolute constraint on the sorting of items in the sorting, and the relative preference represents the constraint on the relative sorting between items in the sorting.

[0142] In one of the embodiments, the preference stubbornness is initially 1, and when it is necessary to execute according to the preference stubbornness, the preference stubbornness decreases after execution.

[0143] For the specific definition of the sorting device based on multi-stage view interaction, reference can be made to the definition of the sorting method based on multi-stage view interaction in the above text, which will not be elaborated here. Each module in the above sorting device based on multi-stage view interaction can be implemented in whole or in part by software, hardware, and their combination. The above modules can be embedded in the processor in the computer device in hardware form or independent of it, or stored in the memory in the computer device in software form, so that the processor can call and execute the operations corresponding to the above modules.

[0144] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structural diagram may be as shown in Figure 3 . The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a sorting method based on multi-stage view interaction. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device may be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0145] Those skilled in the art can understand that Figure 3 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0146] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the method in the above embodiment.

[0147] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in this application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0148] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0149] The above-described embodiments merely represent several implementation manners of this application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can still be made, and these all belong to the protection scope of this application. Therefore, the protection scope of this application should be subject to the appended claims.

Claims

1. A sorting method based on multi-stage view interaction, characterized in that The method includes: Determining a quadruple of a user in a social network and the user's interactive learning objects; wherein, the quadruple includes: familiarity with the item to be sorted, preference, preference stubbornness, and sorting; the interactive learning objects are determined according to the similarity of users in the social network; In the current interaction round, calculating the scoring contribution of the item to be sorted by the user according to a preset learning rate and the sorting, and calculating the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning objects, and the sorting of the interactive learning objects; the learning rate is used to reflect the degree of adoption of the opinions of neighbors by the user during the opinion update process; Updating the sorting according to the scoring contribution and the aggregated scoring contribution to obtain a learning sorting; Updating the preference degree of the learning sorting according to the preference and the preference stubbornness to obtain an updated sorting; When the consistency ratio of the updated sortings of all users reaches a threshold, outputting a consistent sorting result.

2. The method according to claim 1, wherein Determining the user's interactive learning objects includes: Obtain the sorting of the user and the sorting of the user to be evaluated , where M is the total number of items to be sorted; Calculate the sorting separately and the sorting The normalized probability distributions P and Q are as follows: ； ； Among them, ; Calculate the KL divergence of the probability distribution P with respect to the average probability distribution and the KL divergence of the probability distribution Q with respect to the average probability distribution respectively : : ； ； Among them, ; According to the KL divergence and the KL divergence the JS divergence is calculated as follows: ； Converting the JS divergence to a JS distance: ； When the JS distance between the user and the user to be evaluated is less than the threshold, determining the user to be evaluated as an interactive learning object.

3. The method according to claim 1, wherein Calculating the scoring contribution of the item to be sorted by the user according to a preset learning rate and the sorting includes: Calculating the scoring contribution of the item to be sorted by the user according to a preset learning rate and the sorting as: Among them, represents the scoring contribution, represents the position of item i in the said sorting, M is the total number of items in the sorting, represents the learning rate.

4. The method according to claim 3, characterized in that, Calculating the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning objects, and the sorting of the interactive learning objects includes: Calculating the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning objects, and the sorting of the interactive learning objects as: ； Among them, represents the familiarity weight of the interactive learning object , K is the total number of interactive learning objects, represents the familiarity of the interactive learning object . represents the ranking of item i for the interactive learning object .

5. The method according to claim 4, wherein Updating the sorting according to the scoring contribution and the aggregated scoring contribution to obtain a learning sorting includes: Adding the scoring contribution and the aggregated scoring contribution to obtain a comprehensive score as: ； Sorting the items in the sorting in descending order according to the comprehensive score to obtain a learning sorting.

6. The method according to any one of claims 1 to 5, characterized in that, The preference includes: absolute preference and relative preference; wherein, the absolute preference represents an absolute constraint on the sorting of items in the sorting, and the relative preference represents a constraint on the relative sorting between items in the sorting.

7. The method according to any one of claims 1 to 5, characterized in that, The preference stubbornness is initially 1, and when it is necessary to execute according to the preference stubbornness, the preference stubbornness decreases after execution.

8. A sorting device based on multi-stage view interaction, characterized in that, The device includes: An initialization module, configured to determine a quadruple of a user in a social network and the user's interactive learning objects; wherein, the quadruple includes: familiarity with the item to be sorted, preference, preference stubbornness, and sorting; the interactive learning objects are determined according to the similarity of users in the social network; A scoring calculation module, configured to, in the current interaction round, calculate the scoring contribution of the item to be sorted by the user according to a preset learning rate and the sorting, and calculate the aggregated scoring contribution of the set of interactive learning objects according to the preset learning rate, the familiarity of the interactive learning objects, and the sorting of the interactive learning objects; the learning rate is used to reflect the degree of adoption of the opinions of neighbors by the user during the opinion update process; A learning update module, configured to update the sorting according to the scoring contribution and the aggregated scoring contribution to obtain a learning sorting; A preference setting module, configured to perform a preference degree update on the learning sorting according to the preference and the preference stubbornness degree to obtain an updated sorting; A sorting output module, configured to output a consistent sorting result when the consistency ratio of the updated sortings of all users reaches a threshold.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

Images