Adaptive Deployment of Multi-Agent Systems on Cloud Environments

Chahrazed Labba

16/3/2017

Abstract:

Multi-agent systems (MAS) are highly dynamic and require powerful computational resources to be properly executed. However, such resources are not available for all organizations due to many reasons such as budget constraints. Thus to support more flexibility and enjoy greater scalability, it is common today to outsource totally or partially various types of applications to the cloud environments. The main challenge in this regard is to ensure an optimal allocation of cloud resources over-time. For MAS, this challenge is deepened due to irregular workload progress and intensive communications between the agents, which may result in high computing and data transfer costs. Our research work contributes to the deployment of MAS on various cloud environments. It proposes a generic pre-deployment method to ensure a cost-efficient adaptive deployment of agent-based systems on various cloud infrastructures. Firstly, we propose a conceptual framework and its operationalization for recommending partitioning algorithms to efficiently distribute MAS on cloud resources. The framework supports the MAS designers in: (i) determining the MAS types based on a set of defined recurrent criteria; (ii) analysing the appropriateness of the partitioning algorithms based on the types of both MAS and target cloud environment; and (iii) providing guidelines to develop new distributed systems. Secondly, we propose a set of new partitioning algorithms that support the deployment of different MAS types on various cloud environments. The algorithms are made up from the combinations of extended versions of two basic graph-based partitioning algorithms, which are Fiduccia-Mattheyses (FM) and Pairwise Movement Fiduccia-Mattheyses algorithms (PMFM). Unlike the existing partitioning algorithms for MAS, the new algorithms support the cloud specifications, allow allocating low cost resources that fulfil the system requirements, and maintain reduced communications costs. Thirdly, our research work provides a prediction process that estimates and optimizes the deployment costs. The process allows monitoring the changes in the MAS requirements in terms of computational resources and invoking, once required, the set of candidate algorithms to enhance the quality of the partitioning. It uses a cost-deployment model to compare the obtained deployment solutions and then recommend the suitable one. To validate our approach, we develop two case studies including the deployment of a crisis evacuation MAS on a hybrid cloud as well as the deployment of a dynamic business process on a cloud federation. Experimental results highlight the efficiency of our approach and show that deployment costs are sensitive to many factors such as the initial deployment, the used partitioning algorithm and the cloud configuration settings.

PresentationChahrazed_LABBA

9h – Accueil des participants
9h15 – Introduction
9h30 – “The Clarion Call from the Crowd: Mining Social Media in
Disaster”, Valerie Shalin, Département de Psychologie, Wright State
University
10h15 – “Un jeu de construction éclairé”, Joffrey Becker, Laboratoire
d’Anthropologie Sociale, LORIA et Frédéric Verhaegen, APEMAC, Université
de Lorraine
11h00 – “News-sites and adblockers : intermediaries of advertising
self-regulation in the field of journalism”, Vassili Rivron, CERReV,
Université de Caen Basse Normandie et Thibault Cholez, LORIA, Université
de Lorraine
11h45 – Discussion
12h15 – Fin du séminaire

Title: The Clarion Call from the Crowd: Mining Social Media in Disaster
Valerie Shalin (Associate professor, Department of Psychology, Wright
State University)

Abstract: Many researchers are attempting to mine social media content
for actionable information during a disaster. Few researchers consider
the processes that generate this content as a source of insight for
identifying actionable content. Simply counting key words ignores the
social, psycholinguistic, cognitive and perceptual processes that may
inform analysis. In this talk we consider lexical choice of the crowd,
in particular the ratio between members of antonym pairs (e.g.,
some/all, alone/together, stop/start) relative to population base rates
as a signal of need. These signals are not related to sentiment and
generalize across disasters. Because they are crowd-based, they cover
multiple, unanticipated perspectives while eliminating reliance on
individual trust metrics.

Bio: Valerie L. Shalin is an associate professor of psychology at Wright
State University, having earned her PhD in learning, developmental, and
cognitive psychology from the University of Pittsburgh in 1987. Her
research addresses planning and communication processes in coordinated
work and corresponding workplace technology for: space exploration,
medicine and surgery, disaster response, and manual labor. With Andrew
Hampton she was the winner of the 2016 Human Factors Prize awarded by
the Human Factors Society, for the work she will be presenting.

———
Titre : Un jeu de construction éclairé
Joffrey Becker (Anthropologue, Chercheur affilié au LAS, Post-Doc au
LORIA) ; Frédéric Verhaegen (Psychologue, Maitre de conférence,
Université de Lorraine)

Résumé : Le projet PsyPhINe (MSH – Université de Lorraine) cherche à
appréhender les questions posées par l’interaction humain/robot à partir
de l’apport de différentes disciplines comme la psychologie, la
philosophie, l’informatique, les neurosciences, l’intelligence
artificielle, la robotique, la linguistique et l’anthropologie. À partir
de diverses expériences conduites autour d’un prototype de lampe
robotisée, le groupe s’intéresse aux questions touchant à l’attribution
d’intentionnalité, d’intelligence, de cognition et d’émotions envers des
existants naturels ou artificiels. Notre intervention reviendra sur les
enjeux qu’implique une telle perspective interdisciplinaire et
présentera quelques résultats des expériences qui ont été menées jusqu’ici.

Bio:
Frédéric Verhaegen est Maitre de conférence en psychologie et en
psychopathologie de la cognition à l’Université de Lorraine. Ses
recherches visent une modélisation globale de l’interaction en tentant
d’intégrer les aspects du comportement no verbal qui participent tout
autant que le comportement verbal à la régulation intersubjective du
processus de communication.

Joffrey Becker est anthropologue, chercheur affilié au Laboratoire
d’Anthropologie Sociale et post-doctorat au LORIA. Ses recherches
portent sur la robotique et plus particulièrement sur les relations
entre humains et machines. Elles visent aujourd’hui à mieux saisir les
dynamiques qui font des robots des outils expérimentaux, dont l’activité
questionne nos modèles sur un plan à la fois ontologique, relationnel et
social.

——–
Title: News-sites and adblockers : intermediaries of advertising
self-regulation in the field of journalism
Thibault CHOLEZ (Maitre de conférence LORIA-Université de Lorraine /
MADYNES-Inria), Vassili RIVRON (Maitre de conférence CERReV-Université
de Caen Normandie / MADYNES-Inria)

Abstract : The so-called free access to journalistic content on most
press sites imposes counterparties in the form of advertisements and
monitoring devices (cookies, trackers). The rapid development of
technical resources, business practices and users resistance strategies
are shaping original forms of intermediation and regulation of the
sector. Conflicts broke out publicly in 2016 in this two-sided market,
opposing advertising, French press sites and adblockers. They
highlighted the economic impact of the massive use of adblockers and
relaunched the debate on the place of advertising in the economy of free
content. The success of adblockers has — along with audience measures,
market research and other analytics —, constituted a new way of
representing the public, their relationship to editorial content and
advertising, and the collection of personal data. The massive use of
adblockers reveals a strong resistance of web users in the face of what
is perceived as drifts of the deregulation of advertising on online
media. We will show how this success has helped to impose on publishers
and advertisers the notions of “reasoned ads”, or at least of
“acceptable ads”, of which adblockers are trying to set themselves up
as guarantors.