Identifying domain-specific accounts

A system, comprising:

• at least one processor;
• and
• at least one non-transitory computer-readable medium containing instructions that, when executed by the at least one processor, are configured to cause the system to perform operations, the operations comprising:
• identifying a plurality of candidate accounts that are followed by a plurality of seed accounts, the seed accounts representative of accounts identified as including knowledge for a particular domain;
• identifying a subset of candidate accounts of the plurality of candidate accounts based on account statistics, the account statistics comprising, for a given candidate account:
• a ratio between a number of followers from among the seed accounts following the given candidate account and a total number of all followers following the given candidate account;
• clustering each candidate account of the subset of candidate accounts into either a first cluster or a second cluster of candidate accounts, the clustering performed based on the account statistics comprising, for the given candidate account:
• the ratio between the number of followers from among the seed accounts following the given candidate account and the total number of all followers following the given candidate account;
• identifying the first cluster as a non-domain-specific cluster and the second cluster as a domain-specific cluster based on a second cluster mean value of the account statistics of accounts of the second cluster being higher than a first cluster mean value of the account statistics of accounts of the first cluster, the account statistics comprising, for the given candidate account:
• the ratio between the number of followers from among the seed accounts following the given candidate account and the total number of all followers following the given candidate account;
• and
• identifying each account of the domain-specific cluster as a domain-specific account for the particular domain.

One or more non-transitory computer-readable media that include instructions that, when executed by one or more processing systems, are configured to cause the processing system to perform operations, the operations comprising:

• identifying a plurality of candidate accounts that are followed by a plurality of seed accounts, the seed accounts representative of accounts identified as including knowledge for a particular domain;
• identifying a subset of candidate accounts of the plurality of candidate accounts based on account statistics, the account statistics comprising, for a given candidate account:
• a ratio between a number of followers from among the seed accounts following the given candidate account and a total number of all followers following the given candidate account;
• clustering each candidate account of the subset of candidate accounts into either a first cluster or a second cluster of candidate accounts, the clustering performed based on the account statistics comprising, for the given candidate account:
• the ratio between the number of followers from among the seed accounts following the given candidate account and the total number of all followers following the given candidate account;
• identifying the first cluster as a non-domain-specific cluster and the second cluster as a domain-specific cluster based on a second cluster mean value of the account statistics of accounts of the second cluster being higher than a first cluster mean value of the account statistics of accounts of the first cluster, the account statistics comprising, for the given candidate account:
• the ratio between the number of followers from among the seed accounts following the given candidate account and the total number of all followers following the given candidate account;
• and identifying each account of the domain-specific cluster as a domain-specific account for the particular domain.