A hybrid-feedback recommender system for employment websites

Recommender systems have been widely used in many fields, but have been particularly common in the generation of user feedback. For Taiwanese employment websites, recommender systems match job seekers with employers but often fail to satisfy both the job seekers’ preferences and the prospective employers’ requirements. In the process, these websites often waste users’ time. To address this problem, this research proposes a hybrid-feedback recommender system specifically for job seekers in the context of Taiwan’s online employment scene. To identify the correlations between job searches and active job-seeking users, this research harness both implicit and explicit techniques that are based on user portfolios and a new data-mining technique called JSS-Tree. With the system, job seekers can spend less time and get better results than is currently possible on Taiwanese employment websites. Furthermore, through JSS-Tree, the recommender system generates substantially diverse recommendation results for job seekers, who thus have more choices than would normally be available and who can avoid wasting time while searching for desirable jobs on employment websites.