federiconitidi
Joined 9 years ago
Karma
44
Submissions
71
Member since
Mar 20
Joined 9 years ago
Right now the script takes into account mainly the person's job title and the location. Job titles are matched against a library of macro positions, to understand the functional role. Location is normalized and then used to identify offices, countries and regions and group accordingly. These data are then fed into a random forest algo that sketches out a best guess of the lines of reporting.
Hope you guys will find it interesting/ useful and if you have any feedback I'd love to hear!