Hiring through Social Networks: A Natural Language Processing Automation of Conversation

Abstract

Linkedin is used by recruiters in the recruitment industry to find candidates for jobs. Recruiters search for candidates, select potential candidates (prospects) and send job advertisement/invitation messages. The messages sent by recruiters to prospects should receive maximum responses/replies from prospects in order to select quality candidates. For maximum responses, the message must contain some quality features like call-to action, personalization, incentives for connecting, job/company credibility, etc. We used Natural Language Processing techniques to classify the message if it contains quality features. To classify the message, we trained different machine learning and deep learn ing models. In response to the recruiter’s message, the prospect asks some questions about the job, for example the job location, salary, benefits, company information, etc. which the recruiter has to answer manually. In this work, we employed the RASA [68] conversational AI framework to develop natural language understanding and con versation models that allow for automated communication between recruiters and job prospects. This automation frees up the recruiter’s time by eliminating the need for them to individually respond to each inquiry about the job.