Prof. Oded Netzer
Columbia Business School
Using hidden Markov models to identify job seekers from social network data.
LinkedIn is the largest professional social network in the world. One of the key motivations for members to use LinkedIn is professional networking and job seeking. At same time, one of Linkedin’s primary revenue sources comes from targeting job seekers. LinkedIn allows users to research companies, showcase their skills and talents, and help the right people and opportunities find their way to the job seeking individuals. One of the key challenges for LinkedIn is to identify job seekers. Job seeking behavior can be indirectly observed though increased activity on the social networking site, such as profile updates or company searches. In this research we use longitudinal individual-level social network activity data from a large sample of LinkedIn members to identify job seekers. We build a hidden Markov model (HMM) in which the different states correspond to different levels of job seeking, where each state is characterized by a multivariate set of behaviors in the social network site. The model allows us to identify the members’ activities on LinkedIn that most likely reflect their job seeking status. We use the model to predict the likelihood of each member’s job seeking status at any point in time. In addition to the longitudinal social network activity data we leverage a two-wave large-scale survey asking members about their job seeking status. We use the HMM to capture the dynamics in job seeking behavior as well as fuse the survey data with the longitudinal social network activity in a natural way. The results of our study can help guide LinkedIn in better targeting job seeking members.