A location history-aware recommender system for smart retail environments

  • PDF / 1,460,386 Bytes
  • 12 Pages / 595.276 x 790.866 pts Page_size
  • 96 Downloads / 241 Views

DOWNLOAD

REPORT


ORIGINAL ARTICLE

A location history-aware recommender system for smart retail environments Thomas Chatzidimitris 1 & Damianos Gavalas 2 & Vlasios Kasapakis 1,3 & Charalampos Konstantopoulos 3,4 & Damianos Kypriadis 4 & Grammati Pantziou 3,5 & Christos Zaroliagis 3,6 Received: 17 October 2019 / Accepted: 23 January 2020 # Springer-Verlag London Ltd., part of Springer Nature 2020

Abstract Recommender systems (RSs) represent integral parts of e-commerce platforms for almost two decades now. The recent emergence of mobile context-aware RSs (CARS) contributed in improving the relevance of recommendations derived by “traditional” RSs through adapting them to the situational user context. This article presents the design and implementation aspects of a collaborative filtering-based mobile CARS, which has been integrated in a smart retailing platform that enables location-based search for retail products and services. In addition to user location, the introduced CARS considers several context parameters like time, season, demographic data, consumer behavior, and location history of the user in order to derive more meaningful product recommendations. Our RS has undergone field trials as well as formal laboratory evaluation tests demonstrating higher accuracy and relevance of recommendations compared with two baseline approaches. Keywords Recommender system . Collaborative filtering . E-commerce . M-commerce . Retailer shop . Shopping mall . Smart retailing . Location-based search . Context awareness . Location history

1 Introduction Recommendation systems (RSs) are information filtering systems aiming at predicting the “rating” (i.e., the preference) that a user would give to an information item (e.g., music file, book, or any other product) or social element (e.g., people or groups) she has not yet considered [8]. RSs recommend those items predicted to better match user preferences, thereby reducing the user ’s cognitive and information load.

Recommendations are made either implicitly (e.g., through ranking a list of information items or displaying a “those you bought this product, also bought those” bar) or explicitly (when the user requests a recommendation). The applications of RSs include, but are not limited to, the areas of e-commerce, e-learning, e-library, e-government, e-tourism, and ebusiness services [17]. Τhe rapid development of mobile computing technologies generated a new thread of research within the field of

* Damianos Gavalas [email protected]

Christos Zaroliagis [email protected]

Thomas Chatzidimitris [email protected]

1

Department of Cultural Technology & Communication, University of the Aegean, Mytilene, Greece

Vlasios Kasapakis [email protected]

2

Department of Product & Systems Design Engineering, University of the Aegean, Syros, Greece

3

CTI, Patras, Greece

4

Department of Informatics, University of Piraeus, Piraeus, Greece

Damianos Kypriadis [email protected]

5

Department of Informatics & Computer Engineering, University of West Attica, Athens, Greece

Grammati Pantziou pa