Article Preview
Top1. Introduction
Adapting information to specific user needs is increasingly fundamental with the explosion of available data in information systems brought on by the advent of new technologies or services such as social network platforms, social media, the Internet of Things, big data, or cloud computing environments. If there is increasing information available in these systems, accessing these contents is increasingly difficult for users because of the high quantity and diversity of information that may interest them. This leads to information overload (Guo et al. 2020; Li et al. 2012) and a high increase in the user’s cognitive load. Therefore, it is more difficult for the user to quickly find the information corresponding to his specific expectations. To avoid this problem, personalized or adaptive systems have been proposed with the aim of presenting the information corresponding to the user’s specific needs (e.g., recommender systems, adaptive hypermedia, personalized information retrieval, adaptive user interfaces). A wide range of application domains are concerned (both on the Internet and in enterprise information systems), such as e-commerce (e.g., Amazon) (Smith and Linden 2017; Linden et al. 2003), video content (e.g., Youtube, Netflix) (Gomez-Uribe and Hunt 2015, Davidson et al. 2010), search engines (e.g., Google)(Speretta and Gauch 2005), e-learning (Wang and Wang 2021; Fink and Kobsa 2002), virtual reality (Griol et al. 2019), health (Mao et al. 2020; Glykas and Chytas 2004), and tourism (Al Fararni et al. 2021; Fink and Kobsa 2002). User modeling or user profiling is very important and fundamental for all these systems and applications because they all require a good inference of the user’s needs. A user profile (or user model) can be defined as a summary of the user’s interests, characteristics, behaviors, or preferences. In contrast, user profiling (or user modeling) collects, organizes, and infers user profile information. Information in the user profile can be explicitly provided by the user (explicit user profile), or more frequently, analyzed implicitly by using interaction data between the users and the system (implicit user profile) (Gauch et al. 2007). Beyond personalized or adaptive systems, user profiling can also be at the base of behavioral analysis systems for improving decision-making, such as anomaly detection systems (Kwon et al. 2021; Wang et al. 2018), fraud detection systems (Lausen et al. 2020; Zhao et al. 2016), customer scoring systems (Esmeli et al. 2020; Ramkumar et al. 2010), influencer or leader detection systems (Girgin 2021; Primo et al. 2021), and terrorist networks (Tundis and Mühlhäuser 2017; Yadav et al. 2019).