| 主权项 |
1. A computer implemented method, comprising:
receiving an indication of a metric value to be optimized; presenting a user interface including a user interface feature to a group of users of a social networking system; collecting response data from the users of the social networking system experiencing the user interface; determining a plurality of attribute values for a user attribute of the group of users of the social networking system; for each particular attribute value of the plurality of attribute values, calculating a metric value for a subset of the group of users of the social networking system who have the user attribute of that particular attribute value, based on the collected response data; determining a plurality of interface rules based on the calculated metric value for the subset of the group of users of the social networking system who have the user attribute of that particular attribute value; storing the determined plurality of interface rules; generating a correlation map for a machine learning model between user interface features and user attributes by examining the stored plurality of interface rules, and recording in the correlation map frequencies of pairs of user attributes and interface features occurring in the plurality of interface rules, wherein the recording comprises: monitoring the users' interaction with the social networking system, updating the plurality of interface rules based on the users' interaction with the social network system, counting the frequencies of pairs of the user attributes and the interface features occurring in the interface rules that are updated based on the users' interaction with the social networking system; generating a list of frequently occurred user attributes by examining the stored plurality of interface rules; determining at least one interface rule of applying the user interface feature depending on the user attribute, wherein the determining at least one rule of applying the user interface feature comprises: determining at least one interface rule of applying the user interface feature depending on a frequently occurred user attribute from the list of the frequently occurred user attributes based on an analytics using a machine learning model, and wherein the determining of interface rule is based on an analytics using the machine learning model based on the metric values for the subsets of the group of users so that the metric value is optimized, wherein the user attribute is correlated to the user interface feature indicated in the interface rule according to the correlation map for the machine learning model; storing the determined at least one interface rule; generating the user interface according to the determined at least one interface rule; and presenting the generated user interface to a user of the social networking system, wherein the user is associated with the attribute value that matches the attribute value associated with the interface rule. |
