• PDF / 531,252 Bytes
• 6 Pages / 439.37 x 666.142 pts Page_size
• 23 Downloads / 223 Views

DOWNLOAD

REPORT

---

PRINCE research group, University of Sousse, Tunisia [email protected] 2 ISITCom Hammam Sousse, University of Sousse [email protected]

Abstract. With the advent of mobile technology, Wireless Sensor Networks as well as Smart wearable devices, relevant Mobile Affective based learning systems could be developed. This enables not only to recognize users’ spontaneous affect state but also to react appropriately to that state. However, additional complex requirements are arising such as handling affect data heterogeneity as well as continuous and unpredictable dynamic changes of the sensing capabilities. The present work aims therefore to overcome these requirements by providing flexible and adaptive software architecture based on semantic models. Keywords: Affective Intelligent Learning Systems · Adaptive software architecture · Mobile Affective Computing · Semantic affective models.

1

Introduction

Currently, several categories of Intelligent Learning Environments (ILE) such as Intelligent Learning Management Systems (ILMSs) or Tutoring Systems (ITSs) are more and more available. With the advent of Affective Computing (AC) technology, systems are increasingly integrating explicitly features for processing users' emotional and affective states. Video games were first to integrate those features followed by ITSs [12]. However, LMSs have not yet integrated those aspects. Moreover, the AC domain is continuously evolving especially with the emergence of mobile computing, Wireless Sensor Networks and wearable computing. Smart phones are also increasingly enriched with sensing features. Combining Smart mobile wearable devices (e.g. Glasses, Watches), sensors and AC offers new research opportunities and creates what is called Mobile AC [15]. The main question that is raised in the present research work is the following: What kind of software architecture is it possible to use in order to take into account ILE’s and MAC’s advantages and build a single flexible and adaptive Mobile Affect ILE? In order to answer the raised question, the rest of the paper is structured as follows. In section 2 related works are presented, and classified. The section 3 presents fundamental aspects of the proposal. It describes (1) specific requirements and motivating orientations, (2) the process that relates affective © Springer Science+Business Media Singapore 2017 E. Popescu et al. (eds.), Innovations in Smart Learning, Lecture Notes in Educational Technology, DOI 10.1007/978-981-10-2419-1_18

121

122

M. Khemaja and A. Taamallah

states acquisition and analysis to learning decision making, (3) fundamental models on which relies the process. The section 4 presents the proposed software architecture while the section 5 presents a prototype implementing the solution. The experimentation of the prototype is also realized through simulation of affect states acquisition and a concrete example using a MAC headset detecting meditation and attention related to Electroencephalograph (EEG) signals. Finally, a conclusion allows positioning