Method and apparatus for identifying re-subscribed user

摘要

Embodiments of the present invention provide a method for identifying a re-subscribed user and a corresponding apparatus. Based on call records of a user, communication numbers of the user which meet a preset condition are identified. Then a re-subscribed user is identified by calculating a communication number coincidence rate between users within different charging periods. In accordance with the present invention, call records of users are used as a basis of comparative analysis, thereby improving an accuracy rate of identifying a re-subscribed user; a new method is practiced to sequenced communication numbers of a user and group communication records, thereby effectively reducing the number of calculations and improving calculation efficiency.

主权项

1. A method for identifying a re-subscribed user, the method comprising:
performing statistical analysis to generate a communication record for each user within a first charging period, wherein each communication record records communication numbers of an individual user within the first charging period which are sequenced according to a preset rule; performing statistical analysis to generate a communication record for each new user who is within a second charging period but not within the first charging period, wherein each communication record records communication numbers of an individual user within the second charging period which are sequenced according to the preset rule; collecting communication numbers in all communication records generated by means of statistical analysis to form a group number set, and for each group number in the set, searching for at least one communication record that contains the group number among all communication records generated by means of statistical analysis, so as to generate a communication record group corresponding to the group number; for each communication record group that is generated, executing the following:
identifying each communication record in the group,using the communication record as a to-be-compared communication record, andcalculating a coincidence rate between communication numbers in the to-be-compared communication record and communication numbers in each communication record that is within the communication record group and meets a preset condition, wherein the communication record meeting a preset condition belongs to a different charging period than the to-be-compared communication record, and according to the forgoing preset rule, a group number corresponding to the communication record group is ranked first among communication numbers that are contained in both the communication record meeting a preset condition and the to-be-compared communication record; and for each calculated coincidence rate, when the coincidence rate is greater than a preset threshold, concluding that users to which two communication records corresponding to the coincidence rate belong are a same re-subscribed user.