A Comprehensive Survey on Comparisons across Contextual Pre-filtering, Contextual Post-filtering and Contextual Modelling Approaches

Abstract

Recently, there has been growing interest in recommender systems (RS) and particularly in context-aware RS. Methods for generating context-aware recommendations are classified into pre-filtering, post-filtering and contextual modelling approaches. In this paper, we present the several novel approaches of the different variant of each of these three contextualization paradigms and present a complete survey on the state-of-the-art comparisons across them. We then identify the significant challenges that requirebeing addressed by the current RS researchers, which will help academicians and practitioners in comparing these three approaches to select the best alternative according to their strategies.

Holistic User Context-Aware Recommender Algorithm

“…(1) isActive (2) INPUT: u (profile of user) (3) i (item or set of items) (4) OUTPUT: bool (True or False) (5) Begin (6) isActive ⟵ False 7ACTIONS ⟵ {click, search, rate, bought} (8) A ⟵ {x: x is recent action on i by u AND x ∈ ACTIONS} (9) if (A ≠ ∅)): (10) isActive ⟵ True (11) return isActive (12) End ALGORITHM 5: isActive.…”

“…Operations such as this can be run in the background so that they do not disturb user interaction. |I m | (number of items transacted) (4) OUTPUT: S ratio (support ratio of X) (5) Begin (6) count ⟵ 0 (7) for all i in I m (8) if (i �� X) (9) count ⟵ count + 1 (10) S ratio ⟵ (count/|I m |) * 100 (11) return S ratio (12) End ALGORITHM 9: Support.…”

“…Moreover, the incorporated contexts in CARS are not enough for tailoring personalized services to users. One or two contexts are only incorporated in recommender systems, and these are not enough to represent users' interests or preferences [6].…”

Location-Aware Recommender System: A review of Application Domains and Current Developmental Processes

“…Content-based filtering is another category of LARS technique that leverages machine learning algorithms power to measure similarities in features to make recommendations. [12,14,16,23,45] proposed LARS based on CBF techniques, and it differs from the ML technique in the sense that it does not make full use of the algorithms to make the recommendation but rather it calculates item features and compare them to user interests. Simply put, it uses item features to recommend other items similar to what users like.…”