Please use this identifier to cite or link to this item: http://artemis.cslab.ece.ntua.gr:8080/jspui/handle/123456789/17482
Title: Quality of Experience in Cyber-Physical Social Systems: A Cultural Heritage Space Use Case
Authors: Θάνου, Αθηνά
Παπαβασιλείου Συμεών
Keywords: QoE
Quality of Experience
Prospect Theory
Game Theory
Cyber-Physical Social Sysems
Cultural Heritage
Social Systems
Common pool of resources
User behavior modelling
Museum Congestion Management
Pricing
Time management
Museum Touring
QoE modelling
QoE optimization
Issue Date: 7-Nov-2019
Abstract: In this PhD thesis, the focus is placed on the optimization of user Quality of Experience (QoE) in Cyber Physical Social Systems and specifically in cultural heritage spaces. In order to achieve maximization of visitor perceived satisfaction, the challenges associated with visitor optimal decision making regarding touring choices and strategies in a museum or a cultural heritage space are examined and the problem of museum congestion is addressed. Cultural heritage spaces, and museums in particular, constitute a special type of socio-physical system because, in contrast to other social systems like schools or churches, user experience is primarily controlled by the visitors themselves. Such a system also embodies both human behaviors and physical and technical constraints, a fact that makes adopting a socio-technical perspective in order to improve the visiting experience, essential. Within the above setting, quantitative models and functions are initially formulated to express the visitor experience that is gained throughout a touring process. The functions are based on several socio-physical and behavioral factors. Using this QoE modeling approach, the problem of how to optimize visitor route choices is addressed. A social recommendation and personalization framework is also presented that exploits common visitor characteristics and recommends a set of exhibits to be visited. The creation of self-organizing museum visitor communities are proposed as a means to enhance the visiting experience. They exploit visitor personal characteristics and social interactions and are based on a participatory action research (PAR) process. Recommendation Selection and Visiting Time Management (RSVTM) are combined and formulated into a two-stage distributed algorithm, based on game theory and reinforcement learning. In addition, this PhD thesis examines the problem of congestion management in cultural heritage spaces from a more pragmatic perspective, considering visitor behavioral characteristics and risk preferences. The motivation behind this approach arose from the observation that, in cultural heritage spaces, people interact with each other and consequently the decisions and behavior of one visitor influence and are influenced by others. It is, therefore, important to understand the unknown behavior tendencies of visitors especially when making decisions in order to improve their visiting experience and reduce museum congestion. The proposed mechanisms are founded on and powered by the principles of Prospect Theory and the Tragedy of the Commons. Particular attention is paid to modeling and capturing visitor behaviors and decision making under the potential risks and uncertainties which are typically encountered by visitors during their visit. According to their relative popularity and attractiveness, exhibits at a cultural heritage site are classified into two main categories: safe exhibits and Common Pool of Resources (CPR) exhibits. CPR exhibits are considered non-excludable and rivalrous in nature, meaning that they may experience "failure" due to over-exploitation. As a result, a visitor's decision to invest time at a CPR exhibit is regarded as risky because his/her perceived satisfaction greatly depends on the cumulative time spent at it by all visitors. A non-cooperative game among the visitors is formulated and solved in a distributed manner in order to determine the optimal investment time at exhibits for each visitor, while maximizing the visitor's perceived satisfaction. Detailed numerical results are presented, which provide useful insights into visitor behaviors and how these influence visitor perceived satisfaction, as well as museum congestion. Finally, pricing is introduced as an effective mechanism to address the problem of museum congestion. Motivated by several studies that position pricing as a mechanism to prevent overcrowding in museums, this thesis analyzes and studies the impact of different pricing policies on visitor decisions when they act as prospect-theoretic decision-makers. The theory of S-modular games is adopted to determine the time invested by each visitor at exhibits while maximizing satisfaction gained.
URI: http://artemis.cslab.ece.ntua.gr:8080/jspui/handle/123456789/17482
Appears in Collections:Διδακτορικές Διατριβές - Ph.D. Theses

Files in This Item:
File Description SizeFormat 
Thanou_Phd_Thesis.pdfThanou PhD Thesis3.24 MBAdobe PDFView/Open


Items in Artemis are protected by copyright, with all rights reserved, unless otherwise indicated.