Profiling and Personalization in Internet of Things Environments

Introduction

Latest information systems become more and more smart. The systems are aimed at user behavior accumulation, his/her preferences recognition, and adapt their behavior (Horgan, 2018; Zhang & Sundar, 2019). In the Internet of Things environments profiling and personalization plays essential role (Lu, Papagiannidis, & Alamanos, 2018). Such systems aim to connect different devices and services (resources) at any time and any place, giving rise to innovative new applications and services. In this case, it is needed to consider as user profiles that are aimed at accumulated his/her competencies and preferences to use this information during the interaction with another resources as well as resource competence profile that should store information about the resource competencies and constraints that have to be satisfied to enable these competences.

Profiling and personalization in IoT environment are an ability of such environment to filter the content and change it in according to user preferences accumulated in user profile. For acquisition of this preferences to the user profile IoT environment should contain the special service that keeps the history of interaction with user and analyze it. Grouping of the users by preferences provides possibilities to identify users with similar interests and provides the information that is interested for one user to another users in the group based on the context situation. For automation of the interaction between different resources the competence profile of a resource is proposed. This profile keeps information about possibilities of the resource and constraints that have to be satisfied to utilize these possibilities.

In scope of the chapter the information model of the user profile is proposed. This model describes the user interacting with Internet of Things resources. The profile includes the following main sections: “personal information”, “user context”, “competences”, “preferences”, and the “interaction history”. Analysis of existing user profile models showed that the “personal information” section is basic for a user profile and is used for the most systems that support profiling components for storing personal and contact user information. A distinctive feature of the developed user profile model is the accumulation of information about the user context, his/her competencies, preferences, and interaction history as well as the description of this information in an ontologically oriented form. The “user context” section includes information about the user, which changes depending on the current situation. The “user competence” section includes a set of knowledge and skills with defined proficiency level and taking into account the context for application of this knowledge or skill. The section “user preferences” includes information about tasks, methods, settings for presenting results, as well as groups of users formed by preferences. In this case, it is proposed to divide the preferences into explicit and implicit ones. Explicit preferences are indicated by the user, while implicit ones are detected automatically. The “interaction history” section is used to store the user activities for further analysis. The section is used for offline analysis of user actions, splitting users into groups and updating user preferences.

For the proposed information model of the resource profile the following sections were defined: section “basic information”, “competences and constraints”, “resource context”, and “historical information”. The main task of description a resource profile is to formalize the tasks that this resource can perform, and the constraints that have to be met in order for this resource can perform the tasks.