Recommender systems play an important role in electronic commerce (e-commerce). By analyzing the relations that exist between items and users, these systems are able (i) to improve the performance of the mechanisms for searching/discovering information on large-scale spaces, and (ii) to allow commercial platforms to better identify potential target users.

In this paper, the authors focus on job recommender systems, which aim at predicting the job offers that are likely to be of interest for a given user profile. The proposed solution combines the use of two different classic filters, namely collaborative and content-based filters. The authors report their experience at the RecSys Challenge 2016 competition, where their system, despite its relative simplicity, was ranked in the tenth place for its performance.

The abstract looks a bit unusual, mentioning a numeric score related to an unspecified metric, and the keywords that were provided by the authors are probably not the most representative ones. Moreover, the introductory part could have been more comprehensive and better structured, in order to ensure a more self-contained contribution. Last, several references to concepts/solutions mentioned in the paper are missing.