Efficient and Scalable Job Recommender System Using Collaborative Filtering

Abstract

Recommendation system is a techniques, which provides users with information, which he/she may be interested in or accessed in past. Traditional recommender techniques such as content and collaborative filtering used in various applications such as education, social media, marketing, entertainment, e-governance and many more. Content-based and collaborative filtering has many advantages and disadvantage and they are useful in specific application. Sparsity and cold start problem are major challenges in content and collaborative filtering. Challenges of content and collaborative filtering can be solved by using hybrid filtering. Hybrid filtering combines the features of two recommender system like content and collaborative; content-based filtering improves the classification accuracy and collaborative model easily gives the best-predicted result of a latent factor model. In this paper, we have presented a brief survey of the recommendation system approaches, techniques and application, one important application of recommendation system in Job Recruitment; in which candidates are elected by using online job recruitment portal based on their profile and job history and behaviour components; wherein it serves millions of candidates with suitable and personifies jobs. As per the recent survey this domain is less explored till now and existing job recommender system has many shortcomings, they use resumes/profile and job descriptions for analysis and new job post and candidate profiles are not matched properly because of cold start problem, sometime potential candidate loses their job due to the incomplete job description and education detail in the ontology. LinkedIn’s Job Ecosystem handles few problems, few are still unsolved that we discussed in result part. In this paper, we have presented a comparative analysis of different job recommender system and their techniques.