Method for mining social account of target object, server, and storage medium

A server for mining a target social account of a target object, the server comprising:

• a memory storing instructions;
• and
• a processor in communication with the memory, wherein, when the processor executes the instructions, the instructions are configured to cause the server to:
• establish a group community comprising at least two social groups satisfying a preset degree of association relationship by:
• establishing a community network comprising the at least two social groups, wherein:
• the at least two social groups comprise a first node and a second node in the community network, wherein the first node corresponds to a first social group of the at least two social groups and the second node corresponds to a second social group of the at least two social groups,
• in response to the first node and the second node including at least one same social account, data is added as a directed edge between the first node and the second node, and
• an edge weight of the directed edge is equal to a ratio of a first value and a second value, wherein the first value is a number of the at least one same social account in both the first node and the second node, and the second value is a number of social accounts in the second node;
• traversing the community network, using each node as a target node, and collecting a node label of a neighboring node of the target node and an indegree edge weight of a corresponding indegree edge, wherein:
• the target node comprises a node in the community network, and
• an initial value of the node label comprises a node identifier of a corresponding node;
• in response to more than one neighboring nodes having a same node label, adding indegree edge weights corresponding to the more than one neighboring nodes having the same node label, the same node label indicating the more than one neighboring nodes being classified into a same group community;
• replacing a node label of the target node with a node label corresponding to a maximum indegree edge weight;
• and
• after traversing the community network, classifying social groups having the same node label into the group community;
• locate, from the group community, a target social group comprising a target object by using attribute information of the target object;
• parse preset information of a social account in the target social group, and obtain first identity information of a user corresponding to the social account by:
• extracting remark information of each social account in the target social group, the remark information comprising information used to mark an identity of the user, and
• extracting the first identity information from the remark information by:
• identifying a sound representation of the first identity information;
• determine whether the first identity information matches with second identity information of the target object, the first identity information comprising a job title of the user, by:
• determining whether the sound representation of the first identity information matches with a sound representation of the second identity information;
• when it is determined that the first identity information matches the second identity information of the target object, label the social account corresponding to the first identity information as a target social account of the target object;
• determine a target clustering object matching the attribute information of the target object with a matching degree satisfying a preset matching degree, the matching degree indicating association relationship between attribute information of the target clustering object and the target object, and the target clustering object being related to the target social group comprising the target object;
• filter out the group community unrelated to the target clustering subject, and obtaining a target group community related to the target clustering subject, the target group community comprising the target social group;
• and
• label a social group in the target group community as a candidate social group for the target social group;
• and
• wherein the edge weight is calculated according to
• WGy→Gx=Countcom(Gy,Gx)/Count(Gx)
• WGy→Gx represents the edge weight from the first social group Gy to the second social group Gx, Countcom(Gy, Gx) respresents a number of shared social accounts of the first social group Gy and the second social group Gx, and Count(Gx) is a number of social accounts of the first social group.

A non-transitory computer readable storage medium storing instructions, wherein the instructions, when executed by a processor, cause the processor to perform:

• establishing a group community comprising at least two social groups satisfying a preset degree of association relationship by:
• establishing a community network comprising the at least two social groups, wherein:
• the at least two social groups comprise a first node and a second node in the community network, wherein the first node corresponds to a first social group of the at least two social groups and the second node corresponds to a second social group of the at least two social groups,
• in response to the first node and the second node including at least one same social account, data is added as a directed edge between the first node and the second node, and
• an edge weight of the directed edge is equal to a ratio of a first value and a second value, wherein the first value is a number of the at least one same social account in both the first node and the second node, and the second value is a number of social accounts in the second node;
• traversing the community network, using each node as a target node, and collecting a node label of a neighboring node of the target node and an indegree edge weight of a corresponding indegree edge, wherein:
• the target node comprises a node in the community network, and
• an initial value of the node label comprises a node identifier of a corresponding node;
• in response to more than one neighboring nodes having a same node label, adding indegree edge weights corresponding to the more than one neighboring nodes having the same node label, the same node label indicating the more than one neighboring nodes being classified into a same group community;
• replacing a node label of the target node with a node label corresponding to a maximum indegree edge weight;
• and
• after traversing the community network, classifying social groups having the same node label into the group community;
• locating, from the group community, a target social group comprising a target object by using attribute information of the target object;
• parsing preset information of a social account in the target social group, and obtaining first identity information of a user corresponding to the social account by:
• extracting remark information of each social account in the target social group, the remark information comprising information used to mark an identity of the user, and
• extracting the first identity information from the remark information by:
• identifying a sound representation of the first identity information;
• determining whether the first identity information matches with second identity information of the target object, the first identity information comprising a job title of the user, by:
• determining whether the sound representation of the first identity information matches with a sound representation of the second identity information;
• when it is determined that the first identity information matches with the second identity information of the target object, labeling the social account corresponding to the first identity information as a target social account of the target object;
• determining a target clustering object matching the attribute information of the target object with a matching degree satisfying a preset matching degree, the matching degree indicating association relationship between attribute information of the target clustering object and the target object, and the target clustering object being related to the target social group comprising the target object;
• filtering out the group community unrelated to the target clustering subject, and obtaining a target group community related to the target clustering subject, the target group community comprising the target social group;
• labeling a social group in the target group community as a candidate social group for the target social group;
• and
• wherein the edge weight is calculated according to
• WGy→Gx=Countcom(Gy,Gx)/Count(Gx)
• WGy→Gx represents the edge weight from the first social group Gy to the second social group Gx, Countcom(Gy, Gx) respresents a number of shared social accounts of the first social group Gy and the second social group Gx, and Count(Gx) is a number of social accounts of the first social group.