Informatiquehttp://dspace.univ-msila.dz:8080//xmlui/handle/123456789/2532024-03-29T11:54:15Z2024-03-29T11:54:15ZAgent-Based Simulation of Crowd Evacuation Through Complex SpacesMohamed ChatraMustapha Bourahlahttp://dspace.univ-msila.dz:8080//xmlui/handle/123456789/427162024-03-25T11:56:13Z2023-01-01T00:00:00ZAgent-Based Simulation of Crowd Evacuation Through Complex Spaces
Mohamed Chatra; Mustapha Bourahla
In this paper, we have developed a description of an agent-based model for simulating the
evacuation of crowds from complex physical spaces for escaping dangerous situations. The
model describes a physical space containing a set of differently shaped fences, and
obstacles, and an exit door. The pedestrians comprising the crowd and moving in this space
in order to be evacuated are described as intelligent agents with supervised machine learning
using perception-based data to perceive a particular environment differently. The
description of this model is developed with the Python language where its execution
represents its simulation. Before the simulation, the model can be validated using an
animation written with the same language to fix possible problems in the model description.
A model performance evaluation is presented using an analysis of simulation results,
showing that these results are very encouraging
2023-01-01T00:00:00ZFouille de données des comportements réalistes des foules de piétonsMohamed CHATRAhttp://dspace.univ-msila.dz:8080//xmlui/handle/123456789/427152024-03-25T11:53:15Z2024-01-01T00:00:00ZFouille de données des comportements réalistes des foules de piétons
Mohamed CHATRA
Dans cette thèse, nous avons développé une description d'un modèle basé sur des agents pour simuler l'évacuation de foules d'espaces physiques complexes pour échapper à des situations dangereuses. Ce modèle décrit un espace physique contenant un ensemble de clôtures et d'obstacles de formes différentes, ainsi qu'une porte de sortie. Les piétons composant la foule et se déplaçant dans cet espace afin d'être évacués sont décrits comme des agents intelligents dotés d'un apprentissage automatique supervisé utilisant des connaissances extraites à l’aide de fouille de données sur des données synthétiques et qui sont basées sur la perception pour percevoir différemment un environnement particulier. La description de ce modèle est développée avec le langage Python où son exécution représente sa simulation. Avant la simulation, le modèle peut être validé à l'aide d'une animation écrite avec le langage Python et ce pour résoudre d'éventuels problèmes de description du modèle. Une évaluation des performances du modèle est présentée à l'aide d'une analyse des résultats de simulation et cette évaluation montre que ces résultats sont très encourageants
2024-01-01T00:00:00ZSENTIMENT ANALYSIS BASED ON DEEP LEARNINGBENTOUMI, BILALGHANAI, ZAKARYASupervisor: BOUZAROURA, Ahlemhttp://dspace.univ-msila.dz:8080//xmlui/handle/123456789/424672024-02-27T13:30:00Z2021-06-10T00:00:00ZSENTIMENT ANALYSIS BASED ON DEEP LEARNING
BENTOUMI, BILAL; GHANAI, ZAKARYA; Supervisor: BOUZAROURA, Ahlem
Sentiment Analysis and Opinion Mining is an emerging field, in recent years several
research studies have focused on the task of sentiment analysis, especially in the field of
microblogging.
Our work is part of this same research axis, we propose a system of subjective
classification of the opinions of users of the twitter social network on a product or an event in
three categories: positive, negative and neutral.
Our contribution consists in integration of deep learning methods besides Naturel
Language Processing methods.
Our sentiment analysis system is still in the development phase and far from complete.
For improvement, we propose the use of hybrid approaches to have better results, and
possibly expand the use of this work towards other objectives such as trend analysis and
knowledge extraction from social networks.
2021-06-10T00:00:00ZFast approach for link prediction in complex networks based on graph decompositionAbdelhamid SaifFarid NouiouaSamir Akhroufhttp://dspace.univ-msila.dz:8080//xmlui/handle/123456789/422862024-02-11T12:23:06Z2023-01-01T00:00:00ZFast approach for link prediction in complex networks based on graph decomposition
Abdelhamid Saif; Farid Nouioua; Samir Akhrouf
Social networks such as Facebook, Twitter, etc. have dramatically increased in recent years. These databases are huge and
their use is time consuming. In this work, we present an optimal calculation in graph mining for link prediction to reduce
the runtime. For that purpose, we propose a novel approach that operates on the connected components of a network instead
of the whole network. We show that thanks to this decomposition, the results of all link prediction algorithms using local
and path-based similarity measure scan be achieved with much less amount of computations and hence within much shorter
runtime. We show that this gain depends on the distribution of nodes in components and may be captured by the Gini and
the variance measures. We propose a parallel architecture of the link prediction process based on the connected components
decomposition. To validate this architecture, we have carried out an experimental study on a wide range of well-known
datasets. The obtained results clearly confrm the efciency of exploiting the decomposition of the network into connected
components in link prediction
2023-01-01T00:00:00Z