Set error correction optimization method and device and computer storage medium

Abstract

The invention provides a set error correction optimization method and device and a computer storage medium, and the method comprises the steps: obtaining a to-be-optimized set which comprises at least two target objects; a to-be-optimized edge is analyzed from the to-be-optimized set, and the to-be-optimized edge is a low-frequency abnormal edge and / or an isolated edge; obtaining the relation strength of the to-be-optimized edge; and deleting the to-be-optimized edges of which the relation strength is lower than a preset strength threshold value so as to optimize the to-be-optimized set. Through the mode, the error correction capability is improved, and the content in the set is ensured to be accurate.

Description

Technical Field

[0001] The present application relates to the technical field of data mining, and in particular to a set error correction optimization method, device and computer storage medium. Background Art

[0002] The most common set processing method is clustering algorithm, the core of which is to mine the relationship between target objects in the set from the perspective of the closeness of the relationship.

[0003] The general set mining method mainly estimates the relationship strength from the frequency of the target object, which will inevitably introduce non-target objects in the set. When a large number of non-target objects in the set are introduced into the set, it will cause the set partitioning to be fuzzy and the set partitioning to be wrong. Summary of the invention

[0004] In order to solve the above technical problems, the present application proposes a set error correction optimization method, device and computer storage medium.

[0005] To solve the above technical problems, the present application proposes a set error correction optimization method, which includes: obtaining a set to be optimized, wherein the set contains at least two target objects; analyzing the edges to be optimized from the set to be optimized, wherein the edges to be optimized are low-frequency abnormal edges and / or isolated edge edges; obtaining the relationship strength of the edges to be optimized; deleting the edges to be optimized whose relationship strength is lower than a preset strength threshold, so as to optimize the set to be optimized.

[0006] Among them, analyzing the edges to be optimized from the set to be optimized includes: obtaining a first average number of times that two objects in the set to be optimized meet a preset condition; obtaining a second average number of times that two objects in all similar sets meet the preset condition; obtaining the number of times two objects in the set to be optimized meet the preset condition; obtaining a low-frequency deviation rate of the two objects based on the number of times the two objects meet the preset condition, the first average number of times, and the second average number of times; and determining the edge between two objects whose low-frequency deviation rate is lower than a preset deviation rate threshold as the low-frequency abnormal edge.

[0007] Among them, analyzing the edges to be optimized from the set to be optimized includes: obtaining the collective network topology structure of the set to be optimized, wherein the collective network topology structure includes a number of nodes and edges between nodes; determining nodes that have no adjacent nodes or only one adjacent node as isolated edge nodes; determining the edges adjacent to the isolated edge nodes as isolated edge edges.

[0008] Among them, the node that has no adjacent nodes or only one adjacent node is determined as an isolated edge node, including: performing a first edge analysis on the set to be optimized, and determining the node that has no adjacent nodes or only one adjacent node as an isolated edge node; hiding the isolated edge node and the isolated edge in the set network topology structure; continuing to perform a second edge analysis on the set to be optimized until there is no node belonging to the isolated edge node or the number of analysis times is reached.

[0009] Among them, obtaining the relationship strength of the edge to be optimized includes: obtaining the initial relationship strength of the edge to be optimized in the set to be optimized; obtaining the degree of association between the two objects corresponding to the edge to be optimized; and generating the view relationship strength of the edge to be optimized using the degree of association, the initial relationship strength and the reconciliation parameter.

[0010] Among them, obtaining the degree of association between the two objects corresponding to the edge to be optimized includes: obtaining historical data of the two objects corresponding to the edge to be optimized; identifying the degree of association of the historical data through a view behavior recognition algorithm to obtain the degree of association between the two objects corresponding to the edge to be optimized.

[0011] Among them, the set error correction optimization method also includes: obtaining the set type of the set to be optimized; based on the set type, obtaining the content attributes of the set to be optimized; setting the set paradigm according to the set type, wherein the set paradigm includes a content attribute list and constraints of each content; optimizing the content attributes of the set to be optimized according to the set paradigm.

[0012] The collection type and / or the content attribute are / is classified and predicted by using a classification prediction model.

[0013] In order to solve the above technical problems, the present application proposes a collective error correction optimization device, which includes a memory and a processor coupled to the memory;

[0014] The memory is used to store program data, and the processor is used to execute the program data to implement the above-mentioned set error correction optimization method.

[0015] In order to solve the above technical problems, the present application proposes a computer storage medium, which is used to store program data. When the program data is executed by a computer, it is used to implement the above-mentioned set error correction optimization method.

[0016] Compared with the prior art, the beneficial effects of the present application are as follows: the set error correction optimization device obtains a set to be optimized, wherein the set contains at least two target objects; analyzes the edges to be optimized from the set to be optimized, wherein the edges to be optimized are low-frequency abnormal edges and / or isolated edge edges; obtains the relationship strength of the edges to be optimized; deletes the edges to be optimized whose relationship strength is lower than a preset strength threshold, so as to optimize the set to be optimized. Through the above method, the set relationship error correction capability is improved and the accuracy of the set target objects is guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] in:

[0019] Figure 1 It is a flowchart of the first embodiment of the set error correction optimization method provided by the present application;

[0020] Figure 2 It is a flowchart of the second embodiment of the set error correction optimization method provided by the present application;

[0021] Figure 3 It is a flowchart of the third embodiment of the set error correction optimization method provided by the present application;

[0022] Figure 4 Schematic diagram of isolated edge of suspected abnormal set provided by this application;

[0023] Figure 5 This application provides Figure 1 Schematic diagram of the process of the sub-steps of step S13;

[0024] Figure 6 It is a flowchart of the fourth embodiment of the set error correction optimization method provided by the present application;

[0025] Figure 7 It is a schematic diagram of collective fission caused by collective error correction provided by the present application;

[0026] Figure 8 It is a schematic diagram of the collective error correction data visualization system provided by this application;

[0027] Fig. 9 It is a structural schematic diagram of an embodiment of a collective error correction optimization device provided by the present application;

[0028] Fig.10It is a structural diagram of an embodiment of a computer storage medium provided by the present application. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0030] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can, for example, be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] Please refer to Figure 1 , Figure 1 It is a flowchart of the first embodiment of the set error correction optimization method provided by the present application.

[0032] The collective error correction optimization method of the present application is applied to a collective error correction optimization device, wherein the collective error correction optimization device of the present application can be a server, a terminal device, or a system in which a server and a terminal device cooperate with each other. Accordingly, the various parts included in the collective error correction optimization device, such as various units, sub-units, modules, and sub-modules, can all be set in the server, can all be set in the terminal device, or can be set in the server and the terminal device respectively.

[0033] Furthermore, the above-mentioned server can be hardware or software. When the server is hardware, it can be implemented as a distributed server cluster consisting of multiple servers, or it can be implemented as a single server. When the server is software, it can be implemented as multiple software or software modules, such as software or software modules used to provide distributed servers, or it can be implemented as a single software or software module, which is not specifically limited here.

[0034] like Figure 1 As shown, the specific steps are as follows:

[0035] Step S11: Obtain the set to be optimized.

[0036] In an embodiment of the present application, a set refers to a set of data or images aggregated from multiple objects, the set contains at least two target objects, and there are relationship edges between each two target objects. Set error correction refers to adding or removing target objects and relationships in the corresponding set.

[0037] It should be noted that in the embodiments of the present application, the collection is a broad concept. The collection of the present application is not limited to the same target object with similar or identical features, but also includes different target objects with association relationships and common features. For example, the collection can be a collection of pets with the same owner, a collection of books in the same field, or a collection of buildings belonging to the same building complex.

[0038] Since the set in the embodiment of the present application includes different target objects with association relationships and common characteristics, the number of target objects in the set to be optimized in the embodiment of the present application is two or more.

[0039] The set to be optimized includes but is not limited to the target object content contained in the set itself, the label and serial number of the set, the type of each target object in the set, the attribute information of each target object in the set, the relationship and relationship strength between each target object in the set, and the relationship attributes between each target object in the set.

[0040] In the embodiment of the present application, the target object can be any object, including but not limited to people, animals, buildings, etc., and the relationship includes but is not limited to the relationship between people and animals, the relative position relationship between buildings, etc.

[0041] Specifically, the set to be optimized in the present application can be a pre-generated set or a preset set. The pre-formed set is generated based on big data, and can be obtained by clustering according to attribute information such as images and labels in the big data, or by clustering through relative position relationships and preset rules.

[0042] Among them, in the specific embodiments of the present application, the preset rules may be, for example: the probability of the same pet appearing at the same time as the same owner is high, the architectural features of two buildings belonging to the same park are similar and the distance is close, and the books in the same field in the library have similar semantic features or keyword features.

[0043] Step S12: Analyze the edges to be optimized from the set to be optimized.

[0044] The edges to be optimized are low-frequency abnormal edges and / or isolated edge edges.

[0045] The low-frequency abnormal edges are the edges to be optimized in the set to be optimized determined by the big data low-frequency probability estimation method. The isolated edge edges are the edges to be optimized in the set to be optimized determined by the set network abnormal topological structure mining method.

[0046] For specific acquisition methods, please refer to the flow chart of the second embodiment of the set error correction optimization method provided in this application and the flow chart of the third embodiment of the set error correction optimization method provided in this application.

[0047] Specifically, this application proposes an embodiment for determining low-frequency abnormal edges. For details, see Figure 2 , Figure 2 It is a flowchart of the second embodiment of the set error correction optimization method provided by the present application.

[0048] like Figure 2 As shown, the specific steps are as follows:

[0049] Step S21: obtaining a first average number of times that two objects in the set to be optimized meet a preset condition, and obtaining a second average number of times that two objects in all similar sets meet the preset condition.

[0050] The preset conditions include, but are not limited to, the co-occurrence of two target objects, the distance between two target objects being less than a preset threshold, and the two target objects having common features.

[0051] Specifically, the set error correction optimization device is based on the low-frequency probability estimation of big data, and from the perspective of frequency proportion statistics, it can screen sets with low-frequency abnormal relationships.

[0052] Among them, the first average number of times that two objects in the set meet the preset conditions is N s_avg , the second average number of times that every two objects in all sets meet the preset conditions is N avg .

[0053] Step S22: Obtain the number of times that two objects in the set to be optimized meet a preset condition.

[0054] Specifically, the number of times any two objects in the set meet the preset condition is N times.

[0055] Step S23: based on the number of times the two objects meet the preset condition, the first average number of times and the second average number of times, the low-frequency deviation rate of the two target objects is obtained.

[0056] The low frequency deviation rate is defined as:

[0057]

[0058] Step S24: Determine the edge between two objects whose low-frequency deviation rate is lower than a preset deviation rate threshold as a low-frequency abnormal edge.

[0059] When R Δ Below the low frequency deviation rate threshold R tr When , it is considered that there is a low-frequency abnormal edge in the set, and the set is screened as a suspected abnormal set. For example, in a specific embodiment of the present application, the low-frequency deviation rate threshold R tr Can be chosen as -0.5.

[0060] Through the above method, based on the estimation of the probability of low-frequency interaction of big data, potential suspected abnormal sets are screened instead of processing the entire set, thus reducing subsequent computing power overhead.

[0061] Specifically, the present application proposes an embodiment for determining isolated edge nodes. For details, see Figure 3 and Figure 4 , Figure 3 It is a flowchart of the third embodiment of the set error correction optimization method provided by the present application; Figure 4 The present application provides a schematic diagram of isolated edges of a suspected anomaly set.

[0062] like Figure 3 As shown, the specific steps are as follows:

[0063] Step S31: Obtain the network topology structure of the set to be optimized.

[0064] The collective network topology structure includes several nodes and edges between the nodes.

[0065] Specifically, see Figure 4 In the embodiment of the present application, the definition of a set with isolated edges means that there is at least one isolated edge in the set. The definition of isolated edge nodes and isolated edge means that in the set network, the corresponding node has no child nodes or only one child node is an isolated edge node, and its adjacent edge is an isolated edge.

[0066] Step S32: Determine a node that has no adjacent node or has only one adjacent node as an isolated edge node.

[0067] In one embodiment of the present application, a set error correction optimization device performs a first edge analysis on the set to be optimized, and determines nodes that have no adjacent nodes or only one adjacent node as isolated edge nodes; hides isolated edge nodes and isolated edge edges in the set network topology structure; and continues to perform a second edge analysis on the set to be optimized until there are no nodes belonging to isolated edge nodes or the number of analysis times is reached.

[0068] Step S33: determine the edges adjacent to the isolated edge nodes as isolated edge edges.

[0069] Specifically, the collective error correction optimization device determines all edges adjacent to the isolated edge nodes as isolated edge edges.

[0070] Step S34: Determine the edge between two objects whose low-frequency deviation rate is lower than a preset deviation rate threshold as a low-frequency abnormal edge.

[0071] Through the above method, based on the mining of isolated network topology structures, potential suspected abnormal sets are screened instead of processing the entire set, thus reducing subsequent computing power overhead.

[0072] Step S13: Obtain the relationship strength of the edge to be optimized.

[0073] The set error correction optimization device extracts historical data between two target objects associated with the abnormal edge for the screened suspected abnormal set and the corresponding abnormal edge, and loads the view recognition algorithm to assist the set error correction.

[0074] Regarding how to obtain the relationship strength of the edge to be optimized, this application proposes steps S131 to S133 as sub-steps of step S13. For details, please refer to Figure 5 , Figure 5 This application provides Figure 1 Schematic diagram of the flow chart of the sub-steps of step S13.

[0075] like Figure 5 As shown, the specific steps are as follows:

[0076] Step S131: Obtain the initial relationship strength of the edge to be optimized in the set to be optimized.

[0077] Specifically, the initial relationship strength between target objects in the set is defined as S base .

[0078] Step S132: Obtain the degree of association between the two objects corresponding to the edge to be optimized.

[0079] Specifically, in an embodiment of the present application, a set error correction optimization device obtains historical data of two objects corresponding to the edge to be optimized; identifies the correlation of the historical data through a view behavior recognition algorithm, and obtains the degree of correlation between the two objects corresponding to the edge to be optimized.

[0080] The visual behavior recognition algorithm can process image data or video data. The corresponding video behavior algorithm can obtain the correlation degree between the objects with suspected correlation relationship in the corresponding image or video as S. cv .

[0081] Step S133: Generate the view relationship strength of the edge to be optimized by using the association degree, the initial relationship strength and the reconciliation parameter.

[0082] The collection error correction optimization device combines the association degree of the fusion view behavior recognition algorithm to comprehensively obtain the relationship strength between the collection objects as S enance :

[0083] S enance =S base +(δ-S base )×S cv

[0084] In particular, the specific implementation of the view behavior recognition algorithm involved in step S3 is not within the scope of protection of this patent and will not be described in detail.

[0085] Through the above method, a method of fusing the degree of association and the strength of relationship is proposed for suspected abnormal relationships to obtain the comprehensive strength of relationship and improve the error correction capability of the collection content.

[0086] Step S14: Delete the edges to be optimized whose relationship strength is lower than a preset strength threshold, so as to optimize the set to be optimized.

[0087] Specifically, the set error correction optimization device is based on the comprehensive relationship strength threshold S after the fusion view behavior algorithm. tr , determine whether to keep the abnormal edge. δ is the harmonic parameter, the corresponding value range is 0 to 1, and the typical value is 0.5.

[0088] Furthermore, the present application also proposes an embodiment for performing collective error correction based on a specific scenario. For details, please refer to Figure 6 , Figure 6 It is a flowchart of the fourth embodiment of the set error correction optimization method provided by the present application.

[0089] like Figure 6 As shown, the specific steps are as follows:

[0090] Step S41: Obtain the set type of the set to be optimized.

[0091] The collection type and / or content attribute is / are classified and predicted by utilizing a classification prediction model.

[0092] In one embodiment of the present application, a collective error correction optimization device defines a basic network paradigm and mines collective relationships with anomalies through collective network pattern matching.

[0093] It should be noted that in the embodiments of the present application, different types of sets can define reasonable set paradigms based on actual conditions.

[0094] Step S42: Based on the collection type, obtain the content attributes of the collection to be optimized.

[0095] Specifically, objects in a collection are classified. Based on the category, features, geographic location, interaction relationship and physical relationship between the target objects in the collection, deep learning models and other related algorithms are used to classify and predict the collection in a specific scenario. This application uses an algorithm model to classify and predict the content attributes in a collection, and the corresponding algorithm model can use any model, which will not be repeated here.

[0096] Step S43: Set the collection paradigm according to the collection type.

[0097] The set paradigm includes a list of content attributes and constraints on each content attribute.

[0098] In order to clearly express the set paradigm, the following expressions may be used, including but not limited to: the attributes represented by the nodes, the edges between different nodes, and the spins of the nodes are all expressed as constraints.

[0099] Among them, the constraint conditions include but are not limited to location conditions, quantity conditions, association conditions and other conditions.

[0100] Among them, in the specific embodiments of the present application, the constraints can be, for example: in the collection network of architectural style categories, the logo of each building in the collection should be similar; in the collection network of building complexes, the distance between each pair of buildings in the collection should be correlated; in the book classification network, the field to which each book in the collection belongs should be similar; in the collection network of home categories, the color or related style of each home in the collection should be similar, etc.

[0101] Step S44: Optimize the content attributes of the set to be optimized according to the set paradigm.

[0102] Specifically, the collective error correction optimization device removes the edges that do not meet the constraint conditions from the corresponding collective network.

[0103] Through the above methods, specific collection paradigms are defined for collections in different scenarios to improve the collection error correction capability.

[0104] Please continue to see Figure 7-Figure 8 , Figure 7 is a schematic diagram of collective fission caused by collective error correction provided by this application, Figure 8 This is a schematic diagram of the collective error correction data visualization system provided by the present application. The present application proposes a system display of collective error correction results, assisted re-optimization, and a collective error correction optimization device displays the results of collective error correction in a systematic manner, supporting interactive updating and viewing.

[0105] Specifically, see Figure 7After the set error correction optimization completes the set content error correction, the originally closely connected set will directly split into multiple different sets. At this time, the number of sets increases, and the number of corresponding sets decreases, which improves the accuracy of set mining.

[0106] In the embodiment of the present application, a collective error correction data visualization system is designed to facilitate the checking and correction of collective error correction results.

[0107] Among them, the main interface includes three core application interfaces: collection search, collection topology display, and collection relationship detail display.

[0108] The collection topology display interface supports displaying collection topology, including collection target object nodes and collection target object relationships. The remaining collection target object nodes and relationships after collection error correction are displayed in solid lines or obvious styles, and the collection target object nodes and relationships eliminated by collection error correction are displayed in dotted lines or unclear styles.

[0109] The collection relationship supports mouse clicks, and supports operations such as viewing details, removing relationships, and adding relationships. The interface supports the option to display / hide the relationship for correction, and to cancel or save the results of interactive optimization.

[0110] It also includes a relationship detail display interface. When one of the edges is selected for viewing in the collection topology display interface, the corresponding relationship details display the pictures or videos between the two collection object nodes.

[0111] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0112] In order to implement the above-mentioned set error correction optimization method, this application also proposes a set error correction optimization device, please refer to Fig. 9 , Fig. 9 It is a structural diagram of an embodiment of a collective error correction optimization device provided by the present application.

[0113] The collective error correction optimization device 400 of this embodiment includes a processor 41 , a memory 42 , an input / output device 43 , and a bus 44 .

[0114] The processor 41 , the memory 42 , and the input / output device 43 are respectively connected to the bus 44 . The memory 42 stores program data. The processor 41 is used to execute the program data to implement the collective error correction optimization method described in the above embodiment.

[0115] In the embodiment of the present application, the processor 41 may also be referred to as a CPU (Central Processing Unit). The processor 41 may be an integrated circuit chip having the ability to process signals. The processor 41 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or the processor 41 may also be any conventional processor, etc.

[0116] This application also provides a computer storage medium, please continue to refer to Fig.10 , Fig.10 It is a structural diagram of an embodiment of a computer storage medium provided in the present application. The computer storage medium 600 stores a computer program 61. When the computer program 61 is executed by a processor, it is used to implement the set error correction optimization method of the above embodiment.

[0117] When the embodiments of the present application are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

[0118] The above description is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly used in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A set error correction optimization method, characterized in that: The set error correction optimization method comprises: Acquire a set to be optimized, wherein the set contains at least two target objects; Analyzing the edges to be optimized from the set to be optimized, wherein the edges to be optimized are low-frequency abnormal edges and / or isolated edge edges; Obtaining the relationship strength of the edge to be optimized; The edges to be optimized whose relationship strength is lower than a preset strength threshold are deleted to optimize the set to be optimized.

2. The set error correction optimization method according to claim 1, characterized in that: The step of analyzing the edges to be optimized from the set to be optimized includes: Obtaining a first average number of times that any two objects in the set to be optimized meet a preset condition; Obtain the second average number of times that any two objects in all similar sets meet the preset condition; Obtaining the number of times two objects in the set to be optimized meet a preset condition; Based on the number of times the two objects meet a preset condition, the first average number of times, and the second average number of times, obtaining a low-frequency deviation rate of the two objects; The edge between two objects whose low-frequency deviation rate is lower than a preset deviation rate threshold is determined as the low-frequency abnormal edge.

3. The set error correction optimization method according to claim 1, characterized in that: The step of analyzing the edges to be optimized from the set to be optimized includes: Acquire a set network topology structure of the set to be optimized, wherein the set network topology structure includes a plurality of nodes and edges between the nodes; A node that has no adjacent node or only one adjacent node is determined as an isolated edge node; The edges adjacent to the isolated edge nodes are determined as isolated edge edges.

4. The set error correction optimization method according to claim 3, characterized in that: The step of determining a node having no adjacent nodes or only one adjacent node as an isolated edge node includes: Performing a first edge analysis on the set to be optimized, and determining nodes that have no adjacent nodes or only one adjacent node as isolated edge nodes; hiding the isolated edge nodes and the isolated edge edges in the collective network topology; The second edge analysis is continued for the set to be optimized until there are no nodes belonging to the isolated edge nodes or the number of analysis times is reached.

5. The collective error correction optimization method according to claim 1, characterized in that: The obtaining the relationship strength of the edge to be optimized includes: Obtaining the initial relationship strength of the edge to be optimized in the set to be optimized; Obtaining the degree of association between two objects corresponding to the edge to be optimized; The view relationship strength of the edge to be optimized is generated by using the association degree, the initial relationship strength and the reconciliation parameter.

6. The collective error correction optimization method according to claim 5, characterized in that: The obtaining the association degree between the two objects corresponding to the edge to be optimized includes: Obtain historical data of two objects corresponding to the edge to be optimized; The historical data is identified for correlation degree by using a view behavior recognition algorithm to obtain the correlation degree between the two objects corresponding to the edge to be optimized.

7. The collective error correction optimization method according to claim 1, characterized in that: The set error correction optimization method also includes: Obtaining the collection type of the collection to be optimized; Based on the collection type, obtaining content attributes of the collection to be optimized; Setting a collection paradigm according to the collection type, wherein the collection paradigm includes a content attribute list and constraint conditions of each content; The content attributes of the set to be optimized are optimized according to the set paradigm.

8. The set error correction optimization method according to claim 7, characterized in that: The collection type and / or the content attribute are classified and predicted by utilizing a classification prediction model.

9. A collective error correction optimization device, characterized in that: The collective error correction optimization device includes a memory and a processor coupled to the memory; The memory is used to store program data, and the processor is used to execute the program data to implement the collective error correction optimization method as described in any one of claims 1 to 8.

10. A computer storage medium, characterized in that: The computer storage medium is used to store program data, and when the program data is executed by a computer, it is used to implement the collective error correction optimization method as described in any one of claims 1 to 8.

Images