Khaled Ghedira

This paper addresses the job shop scheduling problem including time lag constraints. This is an extension of the job shop scheduling problem with many applications in real production environments, where extra (minimum and maximum) delays can be introduced between successive operations of the same job. It belongs to a category of problems known as NP-hard problem due to large solution space. Biogeography-based optimization is an evolutionary algorithm which is inspired by the migration of species between habitats, recently proposed by Simon in 2008 to optimize hard combinatorial optimization problems. We propose a hybrid biogeography-based optimization (HBBO) algorithm for solving the job shop scheduling problem with additional time lag constraints with minimization of total completion time. In the proposed HBBO, the effective greedy constructive heuristic is adapted to generate the initial population of habitat. Moreover, a local search metaheuristic is investigated in the mutation step in order to ameliorate the solution quality and enhance the diversity of the population. To assess the performance of HBBO, a series of experiments on well-known benchmark instances for job shop scheduling problem with time lag constraints is performed.

Since the early 1980s, a paradigm shift, caused by the work undertaken in the field of cognitive psychology, has occurred. This shift is known as the move from teacher-centered instruction to learner-centered or learning-centered instruction, and emphasizes the importance of building new knowledge on previous ones, interacting with peers, making meaningful and reflective learning and being engaged in his own path to foster learning. This new vision of teaching has created a need for new learning and assessment instruments that are better adapted to these pedagogical realities. In this context, the electronic portfolio or e-portfolio is one of the most versatile and effective tools that have been proposed for this purpose. More specifically, the interest in e-portfolios has grown considerably with the emergence of the competency-based approach and portfolio-based competency assessments. The purpose of this paper is to describe a semantic-based representation of e-portfolios, defined on the basis of official e-portfolio standards and specifications. Moreover, a comparative study of several well-known e-portfolio solutions has been carried out based on different facets, such as functional features, technical and organizational features. The objective is to identify those features that are mostly supported by e-portfolio solution providers and accordingly to gain a fairly accurate idea of the common structure of e-portfolios. In addition, the authors take advantage of an already implemented ontological model describing competency-related characteristics of learners and learning objects and combine it with the e-portfolio ontology, with a view to support a more reliable and authentic competency assessment. The proposed e-portfolio ontology was built following the ontology development methodology Methontology (Fernandez et al., 1997). In addition, it was constructed using the Protégé ontology environment (Protégé, 2007) and was implemented in OWL (Web Ontology Language) (Antoniou and Harmelen, 2004). The proposed e-portfolio ontology provides humans with a shared vocabulary that enables capturing the most important elements in e-portfolios and serves as the basis for the semantic interoperability for machines. The main advantage of the e-portfolio ontology lies in its ability to provide a common and semantically enriched representation of e-portfolio artifacts, thus facilitating the interoperability and exchange of competency evidences between different learning systems and platforms. In addition, capturing the semantics of e-portfolios helps to make them utilizable by intelligent applications.

During Web services selection processes based on the negotiation approaches, systems initially search for services that comply with the users’ functional requirements. Then, based on the retrieved set of functionally similar services, the negotiators start negotiation in order to come up with an agreement about the QoS parameter preferences. Here the problem occurs when the number of services retrieved during the first step is huge. In such case, the performance of the Web service selection process based on the QoS requirements can be degraded. Examining the specifications of all retrieved services is certainly a waste of time. Instead, it is more reasonable to remove all services that are unavailable and under the users’ requirement expectations before the start of the negotiation. To deal with this issue, in this paper we propose a multi-agent based filtering approach that adopts a Web service monitoring method allowing the filtering and the selection of the best candidate Web servi

A swarm of pedestrians is a large group of individual agents that use autonomous coordination via local control laws. Their emergent behaviour allows crowds to have patterns which are not externally planned. Researchers prove that a crowd in emergency situations has a well-determined emergent behaviour. The pedestrian crowd simulation systems in case of emergency must have these emergent behaviours in order to prove its effectiveness. For this purpose we have created the E-PHuNAC model (Evacuation-Personalities' Human's Nature of Autonomous Crowds). Our model takes into account the three main perspectives for understanding the behaviour of crowds in an evacuation. Indeed, we have considered the normative affiliation, social identity and panic. The aim of this work is to demonstrate the effectiveness of the E-PHuNAC model and also prove that the swarm behaviour of pedestrians' agents in this model allows the emergence of these global patterns. The conducted experiments are consistent with reality and can reproduce evacuation phenomena.