Method and system for mapping short term ranking optimization objective to long term engagement

A system having at least one processor for identifying a target metric, the system comprising:

• a short-term behavior metrics measurement unit implemented on the at least one processor and configured to measure at least one first type of metric computed based on a first period, associated with a first length of time, with respect to each of a plurality of users;
• a long-term engagement metrics measurement unit implemented on the at least one processor and configured to measure at least one second type of metric computed based on a second period, associated with a second length of time, with respect to each of the plurality of users, wherein the second length of time is larger than the first length of time;
• a correlation analysis unit implemented on the at least one processor and configured to compute correlations between each of the at least one first type of metric and each of the at least one second type of metric with respect to the plurality of users;
• and
• a target metric identification unit implemented on the at least one processor and configured to identify a target metric from the at least one first type of metric based on the correlations, wherein the target metric is correlated with the at least one second type of metric.

A non-transitory machine-readable medium having information recorded thereon for identifying a target metric, wherein the information, when read by the machine, causes the machine to perform the following:

• measuring at least one first type of metric computed based on a first period, associated with a first length of time, with respect to each of a plurality of users;
• measuring at least one second type of metric computed based on a second period, associated with a second length of time, with respect to each of the plurality of users, wherein the second length of time is larger than the first length of time;
• computing correlations between each of the at least one first type of metric and each of the at least one second type of metric with respect to the plurality of users;
• and
• identifying a target metric from the at least one first type of metric based on the correlations, wherein the target metric is correlated with the at least one second type of metric.