Predicting and recommending skills in the social enterprise

In this paper, we discuss the need for accurate skills assessments of employees in large, global, client-facing enterprises and shortcomings of existing systems for obtaining and managing expertise. We describe enterprise and social data that can be mined to improve skill assessment processes. We propose a matrix completion approach with side information for improved skill assessment prediction and recommendation, and discuss how outputs can improve existing business process use cases and illuminate new ones. Copyright © 2013, Association for the Advancement of Artificial Intelligence. All rights reserved.