In order to build the formula ‘Top applicants + Top company = Success’ to the satisfaction of both the sides, a little more depth is needed in times of digitalisation. The ‘Social Computing’ area uses artificial neural networks to suggest suitable jobs to users. The resulting publication was recently published and is freely accessible in the spirit of ‘Open Science’.
We encounter automated recommendation technologies on an everyday basis: Which product best suits my preferences? What holiday offers are available for the desired holiday destination? We already know these approaches from various online portals. However, people are also increasingly using social networks such as LinkedIn or Xing to attract recruiters, map professional milestones or go in search of jobs. So, it is not surprising that these recommendation systems are also popular in the career sector.
The Social Computing* team looked into this in depth and, in collaboration with Moshbit and the University of California, published a paper on ‘Using autoencoders for session-based job recommendations.’ This is all about matching the right applicant with the right company, also or especially when all the necessary basic information is not available.
The revealing results have been published in a freely accessible format to promote reproducible research.
*The ‘Social Computing’ team, led by Elisabeth Lex, is dedicated to better understanding, predicting and shaping social behavior in complex social networks and systems.