Method, apparatus and device for determining home network and storage medium

Abstract

The application discloses a kind of determination method, device and equipment of family network and storage medium.The method comprises: obtaining the communication characteristics of N first users;At least one intimacy of at least one user pair formed by N first users is determined according to the communication characteristics;At least one intimacy is used to construct social network with N first users as node;In the social network, determine target user pair, wherein the target user pair includes first target user and second target user, and the first target user and the second target user belong to the same family network, all nodes in the family network are members of the same family, and Q target closed loop paths are screened out in the social network according to the target user pair;The nodes in the first target user, the second target user and the Q target closed loop paths are determined as pruned network;And pruning network is cut, to obtain family network.According to the embodiment of the application, the accuracy of family relationship identification can be improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of network structure, and particularly relates to a method and device for determining a family network, an apparatus, and a storage medium. BACKGROUND

[0002] With the continuous development of the mobile network communication industry and the rapid popularization of mobile phones, communication and exchange between people become more and more convenient and close, and these exchanges build the social network of the real society. The social relationship with the family as the core has irreplaceable value in the mobile communication industry, and accurately mining the family social relationship between users plays a crucial role in customer retention, personalized recommendation of family business, and precision marketing of the mobile communication industry.

[0003] In the existing family relationship identification method of an operator, traditional machine learning algorithms and business experience are mainly used to mine the rules of the communication and location behaviors of users with clear relationships, and finally to find similar user groups to realize the division of the family circle, and then to take the family as a unit. Although these methods preliminarily solve the problem of family relationship division, the feature factors of the family relationship have diversity and complexity, and the existing method only realizes family division and simple static analysis, so there is a certain deficiency in the accuracy of identifying the family relationship by using the existing method. SUMMARY

[0004] The embodiments of the present application provide a method, device, apparatus, and storage medium for determining a family network, which can solve the problem of poor accuracy of identifying the family relationship.

[0005] In a first aspect, the embodiments of the present application provide a method for determining a family network, which comprises the following steps.

[0006] Obtaining communication features of N first users, wherein the communication features include identity features of the first users and behavior features of the first users in the communication process, and N is a positive integer;

[0007] Determining at least one closeness of at least one pair of users according to the communication features, wherein each pair of users in the N first users forms a pair of users, and there is a closeness between each pair of users;

[0008] Constructing a social network with the N first users as nodes by using the at least one closeness;

[0009] Determining a target pair of users in the social network, wherein the target pair of users includes a first target user and a second target user, the first target user and the second target user belong to the same family network, and all nodes in the family network are members of the same family;

[0010] determining P closed loop paths in the social network according to the target user pair, and screening Q target closed loop paths from the P closed loop paths, wherein P is a positive integer, Q is a positive integer less than or equal to P, and the distance from each node in each of the P closed loop paths to the target user pair is equal;

[0011] determining the first target user, the second target user, and the nodes in the Q target closed loop paths as a pruning network;

[0012] trimming the pruning network to obtain a family network when the number of nodes in the pruning network is greater than a preset threshold.

[0013] In some embodiments, the screening of the Q target closed loop paths from the P closed loop paths comprises:

[0014] determining the weight value of each edge in each of the P closed loop paths;

[0015] judging whether each of the P closed loop paths is a target closed loop path according to the weight value of each edge in the closed loop path.

[0016] In some embodiments, the judging whether each of the P closed loop paths is a target closed loop path according to the weight value of each edge in the closed loop path comprises:

[0017] calculating the average value of the weight value of each edge in the closed loop path;

[0018] calculating the variance of the weight value of each edge in the closed loop path according to the average value;

[0019] determining a preset first numerical value and a first sum value of the variance;

[0020] determining the quotient value of the average value as the dividend and the first sum value as the divisor as the correlation value of the closed loop path, wherein the correlation value is used to represent the correlation degree between the closed loop path and the target user pair;

[0021] judging whether each of the P closed loop paths is a target closed loop path according to the correlation value.

[0022] In some embodiments, the trimming the pruning network to obtain a family network when the number of nodes in the pruning network is greater than a preset threshold comprises:

[0023] splitting the pruning network into at least one target structure mode when the number of nodes in the pruning network is greater than a preset threshold, wherein the nodes in each of the target structure modes have the same connection structure.

[0024] calculating conductance of the pruned network by using the at least one target structure pattern;

[0025] trimming the pruned network according to the conductance to obtain a home network.

[0026] In some embodiments, the constructing a social network with the N first users as nodes by using the at least one affinity includes:

[0027] determining the N first users as nodes of the social network;

[0028] for any user pair composed of the N first users, connecting two first users in the user pair to form an edge of the social network if an affinity of the user pair is greater than an affinity threshold.

[0029] In some embodiments, the determining at least one affinity of at least one user pair according to the communication features includes:

[0030] for each of the at least one user pair, inputting a communication feature of the user pair into a pre-trained logistic regression model to obtain a predicted probability value;

[0031] determining the predicted probability value as the affinity of the user pair.

[0032] In some embodiments, after the constructing a social network with the N first users as nodes by using the at least one affinity, the method further includes:

[0033] adjusting the social network by using a target algorithm, the target algorithm can include a shortest path algorithm and a common neighbor algorithm.

[0034] In a second aspect, an embodiment of the present application provides a determination apparatus of a home network, the apparatus includes:

[0035] an acquisition module configured to acquire communication features of N first users, wherein the communication features include identity features of the first users and behavior features of the first users in a communication process, and N is a positive integer;

[0036] a first determination module configured to determine at least one affinity of at least one user pair according to the communication features, wherein each two first users in the N first users form a user pair, and there is an affinity between each user pair;

[0037] a construction module configured to construct a social network with the N first users as nodes by using the at least one affinity;

[0038] a second determining module configured to determine a target user pair in the social network, wherein the target user pair comprises a first target user and a second target user, and the first target user and the second target user belong to a same family network, and all nodes in the family network are members of a same family;

[0039] a third determining module configured to determine P closed loop paths in the social network according to the target user pair, and screen Q target closed loop paths from the P closed loop paths, wherein P is a positive integer, Q is a positive integer less than or equal to P, and a distance from each node in each closed loop path to the target user pair is equal;

[0040] a fourth determining module configured to determine the first target user, the second target user, and nodes in the Q target closed loop paths as a pruning network;

[0041] a pruning module configured to prune the pruning network to obtain a family network when a number of nodes in the pruning network is greater than a preset threshold.

[0042] In a third aspect, an embodiment of the present application provides a device for determining a family network, and the device comprises a processor and a memory storing computer program instructions.

[0043] The processor implements the method for determining a family network as above when executing the computer program instructions.

[0044] In a fourth aspect, an embodiment of the present application provides a computer storage medium, and the computer storage medium stores computer program instructions, and the computer program instructions are executed by a processor to implement the method for determining a family network as above.

[0045] In a fifth aspect, an embodiment of the present application provides a computer program product, and the computer program product comprises computer program instructions, and the computer program instructions are executed by a processor to implement the method for determining a family network as above.

[0046] In the present application, a communication feature of a user is quantified as an intimacy between users, and then a social network is constructed based on the users and the intimacy between the users, and a target user pair belonging to a same family is determined in the social network, and then a pruning operation is performed on the social network based on a contact degree of each node in the social network and the target user pair, to obtain a family network, wherein all nodes in the family network are members of a same family. In the present application, each feature of a user affects the intimacy between users, and then a family network is constructed based on the intimacy between the users, so that all communication features between the users and any contact of the communication features affect the construction of the family network, thereby affecting the identification of a family relationship, and thus ensuring the accuracy of the identification of the family relationship. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart illustrating a method for determining a home network according to an embodiment of this application;

[0049] Figure 2 This is a schematic diagram illustrating a method for determining a home network according to another embodiment of this application;

[0050] Figure 3 This is a schematic diagram of the hardware structure of a device for determining a home network according to an embodiment of this application;

[0051] Figure 4 This is a schematic diagram of the structure of a home network determination device provided in an embodiment of this application. Detailed Implementation

[0052] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples of this application.

[0053] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0054] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The embodiments will be described in detail below with reference to the drawings.

[0055] Specifically, to solve the problems in the prior art, the embodiments of the present application provide a method, apparatus and device for determining a home network and a storage medium. First, the method for determining a home network provided by the embodiments of the present application will be introduced.

[0056] Figure 1 A flowchart of the method for determining a home network provided by an embodiment of the present application is shown. The method comprises the following steps:

[0057] S110, communication features of N first users are acquired, wherein the communication features comprise identity features of the first users and behavior features of the first users in a communication process, and N is a positive integer.

[0058] In the embodiment, the communication features are attribute features exhibited by the users in the communication process, and the communication features can at least comprise basic features of the users, behavior features of the users, service features of the users, etc. The basic features of the users can comprise basic data such as names, mobile phone numbers, genders, ages, and home locations of the users; the behavior features of the users can comprise data such as call records, message records, mobile locations, and base station in-out data representing the behavior of the users; and the service features of the users can comprise data of services handled by the users.

[0059] S120, at least one closeness between at least one pair of users is determined according to the communication features, wherein each pair of users in the N first users forms a pair of users, and there is a closeness between each pair of users.

[0060] In the embodiment, each pair of users in the N first users can form a pair of users, and for each pair of users, the similarity between the two first users in the communication features can be determined according to the communication features of the two first users in the pair of users.

[0061] For example, whether the surnames of the two users are consistent, whether the permanent locations are consistent, the proportion of call times during non-working hours in the daytime on weekdays, the call duration at night on weekdays, the total base station coincidence degree in a month, and other communication features having correlations can be determined, and the closeness between each pair of users is determined by the communication features having correlations, and the higher the closeness between the pair of users, the more likely there is a family relationship between the pair of users.

[0062] S130, a social network with the N first users as nodes is constructed by using the at least one closeness.

[0063] In this embodiment, after determining at least one closeness of at least one user pair composed of N first users, the N first users can be taken as nodes, and some nodes can be connected to construct a social network.

[0064] Whether to connect the nodes corresponding to the two first users can be determined according to the closeness between the two first users; for example, if the closeness between the two first users is greater than 0.5, it means that the two first users have a relatively close association, and therefore the nodes corresponding to the two first users can be connected, otherwise the nodes corresponding to the two first users are not connected.

[0065] S140, determining a target user pair in the social network, wherein the target user pair includes a first target user and a second target user, and the first target user and the second target user belong to a same family network, and all nodes in the family network are members of a same family;

[0066] In this embodiment, after the social network is constructed, since each node in the social network corresponds to a first user, and the first users with higher closeness have a connection relationship. Therefore, the family relationship of the users can be identified based on the constructed social network.

[0067] Specifically, the social network can be trimmed in a certain way, and the nodes with higher connectivity are divided together, and the social network is divided into a plurality of sub-networks, and each sub-network is a family network, and each node in the family network is a family member.

[0068] In this embodiment, it can be determined that the target user pair belongs to a family network, and therefore the node where the target user pair is located in the social network can be determined, and the family network corresponding to the target user pair can be determined based on the node where the target user pair is located.

[0069] S150, determining P closed loop paths in the social network according to the target user pair, and screening Q target closed loop paths from the P closed loop paths, wherein P is a positive integer, Q is a positive integer less than or equal to P, and the distance from each node in each closed loop path to the target user pair is equal;

[0070] In this embodiment, P closed loop paths can be first screened in the social network, and then Q target closed loop paths can be further screened from the P closed loop paths, and the nodes included in the Q target closed loop paths have strong connectivity with the node corresponding to the target user pair. Therefore, the nodes included in the Q target closed loop paths and the node corresponding to the target user pair are likely to belong to the same family network.

[0071] In an embodiment, the first target user can be determined as user i, and the second target user can be determined as user j, then starting from user i and user j respectively, all nodes in the social network can be traversed to retrieve a set of nodes with a distance of m from user i, denoted as retrieve a set of nodes with a distance of l from user j, denoted as Then, a set intersection operation is performed on and Thus, a closed loop path node set of k-degree links between user i and user j can be obtained wherein k is the sum of m and l.

[0072] Through the above method, when k takes different values, multiple different closed loop paths can be obtained. In this way, P closed loop paths can be screened in the social network.

[0073] S160, determining the first target user, the second target user, and nodes in the Q target closed loop paths as a pruned network.

[0074] In the embodiment, since the nodes in the Q target closed loop paths and the target users have strong connectivity, the first target user, the second target user, and the nodes in the Q target closed loop paths can be determined as a subnetwork of the social network, i.e., a pruned network, and a family network in which the target user pair is located can be further determined in the pruned network.

[0075] S170, pruning the pruned network to obtain a family network when the number of nodes in the pruned network is greater than a preset threshold.

[0076] In the embodiment, since the number of members in each family is usually within a certain interval, if the number of nodes in the pruned network is greater than the preset threshold, it can be considered that the pruned network includes more than one family, and thus the pruned network needs to be pruned to obtain a family network. For example, the pruned network can be pruned by a social network partitioning algorithm, a spectral clustering algorithm, or a social network trimming algorithm to obtain a family network.

[0077] If the number of nodes in the pruned network is less than or equal to the preset threshold, it can be considered that the number of nodes in the pruned network meets the expectation of the number of members in a family, and thus the pruned network can be directly used as a family network.

[0078] The application quantifies the communication features of users as the intimacy between users, then constructs a social network based on the users and the intimacy between the users, and prunes the social network based on the contact degree of the users in the social network to obtain a family network, so that each node in the family network is a member in the same family. In the application, each feature of the user affects the intimacy between the users, and then the family network is constructed based on the intimacy between the users, so that all communication features between the users and any contact of the communication features affect the construction of the family network, thereby affecting the identification of the family relationship, thereby ensuring the accuracy of the identification of the family relationship.

[0079] As an optional embodiment, the screening of the Q target closed loop paths from the P closed loop paths comprises:

[0080] For each closed loop path in the P closed loop paths, a weight value of each edge in the closed loop path is determined.

[0081] According to the weight value of each edge in the closed loop path, it is determined whether the closed loop path is a target closed loop path.

[0082] In the embodiment, the closed loop path is a path in a social network, which starts from a node, passes through a plurality of nodes and returns to the starting node, and there is no repeated node in the path, forming a loop. Therefore, each closed loop path includes a plurality of nodes and edges connecting the nodes.

[0083] In addition, in the closed loop path, each edge corresponds to a weight value. The greater the weight value of the edge, the closer the contact between the two nodes connected by the edge, that is, the stronger the connectivity, and the closer the contact between the nodes, the more likely it is a family relationship. Therefore, for each closed loop path, the weight value of each edge in the closed loop path can be calculated, and whether the closed loop path is a target closed loop path can be analyzed based on the weight values.

[0084] As an optional embodiment, according to the weight value of each edge in the closed loop path, it is determined whether the closed loop path is a target closed loop path, comprising:

[0085] The average value of the weight values of each edge in the closed loop path is calculated.

[0086] The variance of the weight values of each edge in the closed loop path is calculated according to the average value.

[0087] A first sum value of a preset first numerical value and the variance is determined.

[0088] determine a quotient value by taking the average value as the dividend and the first sum value as the divisor, and determine the quotient value as an association value of the closed loop path, wherein the association value is used to represent the association degree of the closed loop path and the target user pair;

[0089] determine whether the closed loop path is a target closed loop path according to the association value.

[0090] In the embodiment, for each closed loop path, the weight value of an edge in the closed loop path represents the closeness of each node in the closed loop path, and therefore the connectivity in the closed loop path can be represented by the average value of the weight value of each edge in the closed loop path, and the discrete degree of the weight value can be represented by the variance of the weight value of each edge in the closed loop path, and therefore the association value calculated by using the average value and the variance can be used to represent the association degree of the closed loop path and the target user pair, and the closed loop path with a higher association value is determined as the target closed loop path.

[0091] For example, the first target user is determined as user i and the second target user is determined as user j, and then the closed loop path node set of the k-degree link between user i and user j is obtained by taking user i and user j as the starting nodes respectively. Since k can be any constant, the family relationship link closed loop path between user i and user j is more than one, and therefore we define the closed loop path as a whole set of:

[0092] For v l and v l+1 , 1≤l≤k-1, there are

[0093] For each closed loop path , the average value E sijk and the variance D sijk of all weight values in are counted, and the association value is calculated by using each group of average value and target value:

[0094]

[0095]

[0096]

[0097] wherein i represents user i, j represents user j, wij represents the weight value of an edge in the closed loop path, Aij represents the number of edges in the closed loop path, E sijk represents the average value of the weight value in the closed loop path, and D sijk represents the variance of the weight value in the closed loop path, EDk This indicates the associated value in the closed-loop path.

[0098] Optionally, after determining the correlation value of each closed-loop path, the correlation values ​​can be sorted and a second value can be set, and the closed-loop path with the second-highest correlation value can be determined as the target closed-loop path; alternatively, a correlation value threshold can be set, and the closed-loop path with a correlation value greater than the correlation value threshold can be determined as the target closed-loop path.

[0099] Furthermore, as another implementation, since k can theoretically be any large natural number, but the family communities of users i and j cannot be arbitrarily expanded, we impose a constraint on the k-degree closed loop for users i and j. We compare the correlation values ​​of the k-degree closed loop and the k+1-degree closed loop, and filter the target closed loop path based on the comparison results. That is, the requirement is:

[0100] ED k+1 >ED k +α

[0101] Among them, ED k ED represents the correlation value in a k-degree closed-loop path. k +1 represents the association value in the (k+1) degree closed loop path, and α is a constant that controls the number of k degree closed loops. Generally, α∈[0.01,0.1]. In this scenario, α = 0.05 can be taken.

[0102] In other words, only k-degree target closed-loop paths that meet the above conditions are determined as the final target closed-loop paths, while target closed-loop paths that do not meet the above conditions are filtered out.

[0103] This embodiment uses the k-degree closed-loop path method to mine the similarity between nodes, thereby mining the features at different levels between nodes, and thus accurately constructing a network with strong similarity.

[0104] As an optional embodiment, if the number of nodes in the pruned network exceeds a preset threshold, the pruned network is pruned to obtain a home network, including:

[0105] If the number of nodes in the pruned network is greater than a preset threshold, the pruned network is split into at least one target structure pattern, wherein the nodes in each target structure pattern have the same connection structure.

[0106] The conductivity of the pruned network is calculated using the at least one target structural pattern;

[0107] The pruned network is trimmed according to the conductivity to obtain a home network.

[0108] In this embodiment, the connection direction between user group nodes can be determined based on the call situation between each user group. In this way, every three or more nodes and the edges between them in the pruned network can form a structural pattern, i.e., a motif. A target structural pattern can be selected from multiple structural patterns, and then the pruned network can be split according to the target structural pattern.

[0109] Then, the conductivity of the pruned network is calculated using at least one target structural pattern. Specifically, each motif can be abstracted into an equivalent circuit model to facilitate calculation. Finally, the conductivity of each motif is weighted and combined to obtain the conductivity of the entire pruned network.

[0110] Since electrical conductivity represents the ability to transmit information in a network, in the aforementioned home network, electrical conductivity can represent the efficiency and speed of information or energy transmission between family members, i.e., the degree of connection between home users. Therefore, the pruned network can be pruned based on electrical conductivity to obtain the home network.

[0111] For example, such as Figure 2 As shown in Figure X, there are several common motifs consisting of three nodes and the edges between them. Since the M5 type has the weakest requirement for the connectivity (number of paths) of the three edges compared to other motif types, M5 can be selected as the target structural pattern, splitting the pruned network into at least one target structural pattern. Then, the pruned network is defined as a set of nodes S. Let cutM(S) represent the number of target structural patterns with only some endpoints in S (i.e., the number of cut edges), and volM(S) represent the number of motif endpoints in S (i.e., the number of edges completely contained within the local community S). For the local community S, its conductivity can be determined as:

[0112]

[0113] Therefore, we can easily find groups with high conductivity and segmentation points by establishing a target structure pattern of a 3-person family and finding stable and convergent segmentation units through iterative search, thereby completing the pruning and segmentation of larger family circles (such as more than 10 people).

[0114] As an optional embodiment, the construction of a social network with the N first users as nodes using the at least one level of intimacy includes:

[0115] The N first users are identified as nodes of the social network;

[0116] For any user pair formed by the N first users, if the corresponding closeness of the user pair is greater than a closeness threshold, connecting two first users in the user pair to form an edge of the social network.

[0117] In this embodiment, users are taken as nodes of the social network, and different users are connected by using the closeness between users to form edges of the social network. In this way, the connection and characteristics between users and the pattern of the relationship between users can be revealed by using the network structure, and analysis can be performed thereon to better find users with family relationship.

[0118] As an optional embodiment, the determining of the at least one closeness of at least one user pair according to the communication features comprises:

[0119] For each of the at least one user pair, inputting the communication features of the user pair into a pre-trained logistic regression model to obtain a predicted probability value;

[0120] Determining the predicted probability value as the closeness of the user pair.

[0121] In this embodiment, the closeness between two users in a user pair can be determined by using the trained logistic regression model. Specifically, the communication features of a plurality of user pairs can be obtained in advance, and the above plurality of user pairs are labeled to obtain positive samples and negative samples. Then, the logistic regression model is trained by using the positive samples and the negative samples to obtain the trained logistic regression model.

[0122] The trained logistic regression model can judge the influence of each communication feature on the closeness of the user pair. Therefore, by inputting the communication features of each user pair into the trained logistic regression model, the output result of the logistic regression model, i.e., a predicted probability value, can be obtained by maximum likelihood estimation. The predicted probability value can be used to determine the closeness of the user pair.

[0123] As an optional embodiment, the communication features can include the features in Table 1 below:

[0124] Table 1

[0125]

[0126]

[0127] As an optional embodiment, after the social network with the N first users as nodes is constructed by using the at least one closeness, the method comprises:

[0128] Adjusting the social network by using a target algorithm. The target algorithm can include a shortest path algorithm and a common neighbor algorithm.

[0129] In the embodiment, the shortest path algorithm is an algorithm for calculating the shortest path in a network, which can find the shortest path between two nodes. In the social network described above, if the shortest path between two nodes is short, it means that the connection between the two nodes is close, and they may belong to the same family.

[0130] The common neighbor algorithm is an algorithm based on the similarity of node neighbors, which can calculate the common measure between the neighbor nodes of two nodes. In the social network described above, if the number of common neighbors between two nodes is large, it means that the connection between the two nodes is close, and they may belong to the same family.

[0131] Therefore, after the social network is constructed, the two algorithms described above can delete the nodes that have a particularly loose connection with other nodes in the social network based on the constructed social network, so as to better construct the family network in the subsequent steps.

[0132] Based on the method for determining the family network provided in the above embodiment, the application also provides a specific implementation of the device for determining the family network. Please see the following embodiments.

[0133] First, see Figure 3 The device 300 for determining the family network provided in the embodiments of the application includes the following modules:

[0134] The acquisition module 301 is configured to acquire the communication characteristics of N first users, wherein the communication characteristics include the identity characteristics of the first users and the behavior characteristics of the first users in the communication process, and N is a positive integer;

[0135] The first determination module 302 is configured to determine at least one closeness of at least one user pair according to the communication characteristics, wherein each two first users in the N first users form a user pair, and each user pair has a closeness;

[0136] The construction module 303 is configured to construct a social network with the N first users as nodes by using the at least one closeness;

[0137] The second determination module 304 is configured to determine a target user pair in the social network, wherein the target user pair includes a first target user and a second target user, the first target user and the second target user belong to the same family network, and all nodes in the family network are members of the same family;

[0138] The third determining module 305 is configured to determine P closed loop paths in the social network according to the target user pair, and screen Q target closed loop paths from the P closed loop paths, wherein P is a positive integer, Q is a positive integer less than or equal to P, and the distance from each node in each of the P closed loop paths to the target user pair is equal.

[0139] The fourth determining module 306 is configured to determine the first target user, the second target user, and the nodes in the Q target closed loop paths as a pruned network.

[0140] The pruning module 307 is configured to prune the pruned network to obtain a family network when the number of nodes in the pruned network is greater than a preset threshold.

[0141] As an implementation form of the present application, the third determining module 305 can be further configured to:

[0142] For each of the P closed loop paths, determine the weight value of each edge in the closed loop path.

[0143] According to the weight value of each edge in the closed loop path, determine whether the closed loop path is a target closed loop path.

[0144] As an implementation form of the present application, the third determining module 305 can be further configured to:

[0145] Calculate the average value of the weight value of each edge in the closed loop path.

[0146] According to the average value, calculate the variance of the weight value of each edge in the closed loop path.

[0147] Determine a preset first value and a first sum value of the variance.

[0148] Determine the quotient value of the average value as the dividend and the first sum value as the divisor as the correlation value of the closed loop path, wherein the correlation value is used to represent the correlation degree between the closed loop path and the target user pair.

[0149] According to the correlation value, determine whether the closed loop path is a target closed loop path.

[0150] As an implementation form of the present application, the pruning module 307 can be further configured to:

[0151] When the number of nodes in the pruned network is greater than a preset threshold, split the pruned network into at least one target structure mode, wherein the nodes in each target structure mode have the same connection structure.

[0152] calculate conductivity of the pruned network by using the at least one target structure pattern;

[0153] trim the pruned network according to the conductivity to obtain the family network.

[0154] As an implementation form of the present application, the construction module 303 can be further configured to:

[0155] determine the N first users as nodes of the social network;

[0156] in a case where the corresponding closeness of any user pair composed of the N first users is greater than a closeness threshold, connect two first users in the user pair to form an edge of the social network.

[0157] As an implementation form of the present application, the first determination module 302 can be further configured to:

[0158] for each user pair in the at least one user pair, input the communication feature of the user pair into a pre-trained logistic regression model to obtain a predicted probability value;

[0159] determine the predicted probability value as the closeness of the user pair.

[0160] As an implementation form of the present application, the family network determination apparatus 300 can be further configured to:

[0161] adjust the social network by using a target algorithm, which can include a shortest path algorithm and a common neighbor algorithm.

[0162] The family network determination apparatus provided by the embodiments of the present application can implement each step in the method embodiments described above, and thus repeated descriptions are omitted here.

[0163] Figure 4 A hardware structure schematic diagram of a family network determination device provided by an embodiment of the present application is shown.

[0164] The family network determination device can include a processor 401 and a memory 402 having computer program instructions stored therein.

[0165] Specifically, the processor 401 can include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0166] The memory 402 can include mass storage for data or instructions. As an example and not by way of limitation, the memory 402 can include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disc (e.g., a compact disc (CD) or a digital versatile disc (DVD)), a solid-state drive (SSD), a USB drive, or a combination of two or more of these. Where appropriate, the memory 402 can include removable or non-removable (or fixed) media. Where appropriate, the memory 402 can be internal or external to the integrated gateway disaster recovery device. In particular embodiments, the memory 402 is non-volatile, solid-state memory.

[0167] The memory can include read-only memory (ROM), random-access memory (RAM), magnetic disk storage mediums, optical storage mediums, flash memory devices, electrical, optical, or other physical / tangible memory storage devices. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software that, when executed (by one or more processors), is operable to

[0168] The processor 401 implements the determination method of the home network in any one of the above embodiments by reading and executing computer program instructions stored in the memory 402.

[0169] In one example, the determination device of the home network can further include a communication interface 403 and a bus 410. Wherein, as shown in the figure, the processor 401, the memory 402, the communication interface 403 are connected through the bus 410 and complete the communication between each other. Figure 4

[0170] The communication interface 403 is mainly used to realize the communication between each module, device, unit and / or equipment in the embodiments of the present application.

[0171] ​Bus 410 includes hardware, software, or both, to couple components of the home network's determined device to each other and to couple components to other components within the home network. While bus 410 is shown for the sake of clarity as a single bus, it can comprise one or more buses operating together. Bus 410 can be implemented using any suitable type of bus or buses, including, but not limited to, an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association local (VLB) bus, or any other suitable bus or interconnect, or a combination of two or more of these. Where appropriate, bus 410 can be used to implement more than one bus over multiple time periods. Although the example embodiments described and illustrated herein relate to a particular bus configuration, the present application contemplates any suitable bus configuration or interconnect.

[0172] The home network's determined device can be based on the above embodiments, thereby realizing the home network's determination method and device as described above.

[0173] In addition, in combination with the home network's determination method in the above embodiments, the embodiments of the present application can provide a computer storage medium to realize. The computer storage medium has computer program instructions stored thereon; the computer program instructions are executed by a processor to realize any one of the home network's determination methods in the above embodiments, and can achieve the same technical effects, to avoid repetition, which will not be described here. Among them, the above computer readable storage medium can include a non-transitory computer readable storage medium, such as a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, etc., which is not limited here.

[0174] In addition, the embodiments of the present application also provide a computer program product, including computer program instructions, which can realize the steps and corresponding contents of the above method embodiments when executed by a processor.

[0175] It needs to be clear that the present application is not limited to the specific configurations and processes described above and shown in the drawings. For the sake of brevity, detailed descriptions of well-known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between steps, after understanding the spirit of the present application.

[0176] The functional blocks shown in the structural block diagrams above can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, and the like. When implemented in software, the elements of the present application are program or code segments that are used to perform the required tasks. The program or code segments can be stored in a machine-readable medium, or transmitted through a data signal carried in a carrier wave over a transmission medium or communication link. A "machine-readable medium" includes any medium that can store or transport information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memory, erasable ROMs (EROMs), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, and the like. The code segments can be downloaded via computer networks such as the Internet, intranets, and the like.

[0177] It is also important to note that the examples in the present application are described based on a series of steps or apparatuses to describe some methods or systems. However, the present application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the examples, or in an order different from the examples, or several steps can be performed simultaneously.

[0178] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0179] The above is merely specific implementation of the present application, and those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, modules and units can refer to the corresponding processes in the foregoing method embodiments, which will not be described herein. It should be understood that the protection scope of the present application is not limited thereto, and any skilled person in the art can easily think of various equivalent modifications or replacements within the technical range disclosed in the present application, and these modifications or replacements should be covered within the protection scope of the present application.

Claims

1. A method for determining a home network, characterized in that, The method includes: Obtain N communication characteristics of the first user, wherein the communication characteristics include the identity characteristics of the first user and the behavioral characteristics of the first user during the communication process, and N is a positive integer; At least one degree of intimacy is determined for at least one user pair based on the communication characteristics, wherein every two first users among the N first users form a user pair, and there is an intimacy between each user pair; A social network is constructed using the at least one level of intimacy, with the N first users as nodes; In the social network, target user pairs are identified, wherein the target user pairs include a first target user and a second target user, the first target user and the second target user belong to the same family network, and all nodes in the family network are members of the same family; Based on the target user pair, P closed-loop paths are determined in the social network, and Q target closed-loop paths are selected from the P closed-loop paths, where P is a positive integer, Q is a positive integer less than or equal to P, and the distance from each node in each of the P closed-loop paths to the target user pair is equal. The first target user, the second target user, and the nodes in the Q target closed-loop paths are defined as a pruned network; If the number of nodes in the pruned network exceeds a preset threshold, the pruned network is pruned to obtain a home network.

2. The method for determining a home network according to claim 1, characterized in that, The step of selecting Q target closed-loop paths from the P closed-loop paths includes: For each of the P closed-loop paths, determine the weight value of each edge in the closed-loop path; Based on the weight value of each edge in the closed-loop path, determine whether the closed-loop path is the target closed-loop path.

3. The method for determining a home network according to claim 2, characterized in that, The step of determining whether the closed-loop path is the target closed-loop path based on the weight value of each edge in the closed-loop path includes: Calculate the average weight value of each edge in the closed-loop path; Calculate the variance of the weight value of each edge in the closed-loop path based on the average value; Determine the preset first value and the first sum of the variances; The quotient of the first sum as the divisor, with the average value as the dividend, is determined as the correlation value of the closed-loop path, wherein the correlation value is used to characterize the degree of correlation between the closed-loop path and the target user pair. The closed-loop path is determined based on the correlation value to determine whether it is the target closed-loop path.

4. The method for determining a home network according to claim 1, characterized in that, When the number of nodes in the pruned network exceeds a preset threshold, the pruned network is pruned to obtain a home network, including: If the number of nodes in the pruned network is greater than a preset threshold, the pruned network is split into at least one target structure pattern, wherein the nodes in each target structure pattern have the same connection structure. The conductivity of the pruned network is calculated using the at least one target structural pattern; The pruned network is trimmed according to the conductivity to obtain a home network.

5. The method for determining a home network according to claim 1, characterized in that, The construction of a social network with the N first users as nodes using the at least one level of intimacy includes: The N first users are identified as nodes of the social network; For any user pair consisting of the N first users, if the intimacy level of the user pair is greater than the intimacy threshold, the two first users in the user pair are connected to form the edge of the social network.

6. The method for determining a home network according to claim 1, characterized in that, Determining at least one level of intimacy between at least one user pair based on the communication characteristics includes: For each user pair in the at least one user pair, the communication characteristics of the user pair are input into a pre-trained logistic regression model to obtain a predicted probability value; The predicted probability value is determined as the intimacy level of the user pair.

7. The method for determining a home network according to claim 1, characterized in that, After constructing the social network with the N first users as nodes using the at least one affinity, the method further includes: The social network is adjusted using a target algorithm, which may include a shortest path algorithm and a common neighbor algorithm.

8. A device for determining a home network, characterized in that, The device includes: The acquisition module is used to acquire N communication characteristics of the first user, wherein the communication characteristics include the identity characteristics of the first user and the behavioral characteristics of the first user during the communication process, and N is a positive integer; The first determining module is used to determine at least one degree of intimacy for at least one user pair based on the communication characteristics, wherein each pair of N first users forms a user pair and there is an intimacy between each user pair; A building module is used to construct a social network with the N first users as nodes using the at least one affinity; The second determining module is used to determine a target user pair in the social network, wherein the target user pair includes a first target user and a second target user, the first target user and the second target user belong to the same family network, and all nodes in the family network are members of the same family; The third determining module is used to determine P closed-loop paths in the social network based on the target user pair, and to filter out Q target closed-loop paths from the P closed-loop paths, wherein P is a positive integer, Q is a positive integer less than or equal to P, and the distance from each node in each of the P closed-loop paths to the target user pair is equal. The fourth determining module is used to determine the first target user, the second target user, and the nodes in the Q target closed-loop paths as a pruned network; The pruning module is used to prune the pruning network to obtain a home network when the number of nodes in the pruning network exceeds a preset threshold.

9. A device for determining a home network, characterized in that, The devices defining the home network include: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the method for determining a home network as described in any one of claims 1-7.

10. A computer storage medium, characterized in that, The computer storage medium stores computer program instructions, which, when executed by a processor, implement the method for determining a home network as described in any one of claims 1-7.

Images