Jean-Emile Symphor

Université des Antilles | Reunion · Department of Geography

Background In many geographic areas influenced by tropical and temperate climates, natural forest ecosystems have been destroyed in favour of plantations of allochthonous trees which are economically profitable for different aspects of the timber industry. Some of these mature plantations degrade the soils and inhibit the regenerations of local flo...

Martinique was recognised as a pioneer in the insular Caribbean for coffee cultivation in the early 18th century. The island is known for its legendary "Martinique" coffee dating from that time but it no longer has coffee-growing industries. As part of a project to revive this cultivation, a geographic information system (GIS) was developed in orde...

Abstract: At planetary as well as local scale, the relief forms induce spatial modifications of the structure of the main bioclimatic factors such as temperature, rainfall, humidity, insolation, evaporation and nebulosity. Depending on the area under consideration, these changes result in more or less strong floristic gradients. We wanted to study...

The complexity of the Lesser Antilles vegetation must be linked to the variety of topographic features that influence the structure of some climatic parameters, particularly the rainfall. The above mentioned environmental factors influence a plurality of biotopes colonized by specific phytocenoses. Using survey transects of different minimal areas,...

Throughout this paper, we will present a new approach that allows the precocious products trends detection within the framework of commercial activities of themass-market retailing. As regards a field in which the competition is very vivacious between the different shop signs and which involves colossal financial wagers, commercial strategies have...

In this paper, we present an heuristic based approach for feature selection in the context of flood prediction. Features are complex variables that represent aggregate values. We apply a preprocessing method on data in order to elicit relevant information that could not be easily accessible initially because it is split through several lines of a d...

Les entrepôts de données occupent aujourd'hui une place centrale dans le processus décisionnel. Outre leur consultation, une des finalités des entrepôts est de servir de socle aux techniques de fouilles de données. Malheureusement, les approches existantes exploitent peu les particularités des entrepôts (multidimensionnalité, hiérarchies et données...

Mining sequential patterns in data streams is a new challenging problem for the datamining community since data arrives sequentially in the form of continuous rapid and infinite streams. In this paper, we propose a new on-line algorithm, SPAMS, to deal with the sequential patterns mining problem in data streams. This algorithm uses an automaton-bas...

Nous présentons dans cet article un nouvel automate : le FIA qui permet de traiter de façon efficace la problématique de l'extraction des itemsets fréquents dans les data streams. Le FIA est une structure de données très compacte et informative qui, par ailleurs, présente des propriétés incrémentales facilitant grandement la mise à jour avec une gr...

Nous présentons dans cet article un nouvel algorithme permettant la construction et la mise à jour incrémentale du FIAθ : FIASCO. Notre algorithme effectue un seul passage sur les données et permet de prendre en compte les nouveaux batches, itemset par itemset et pour chaque itemset, item par item.

Given a set of strings, the Common Subsequence Automaton accepts all common subsequences of these strings. Such an automaton can be deduced from other automata like the Directed Acyclic Subsequence Graph or the Subsequence Automaton. In this paper, we introduce some new issues in text algorithm on the basis of Common Subsequences related problems....

Recently the knowledge extraction community takes a closer look to new models where data arrive in timely manner like a fast and continous o w, i:e: data streams. As only a part of the stream can be stored, mining data streams for sequential patterns and updating pre- viously found frequent patterns need to cope with uncertainty. In this paper, we...

A data stream is a potentially uninterrupted flow of data. Mining this flow makes it necessary to cope with uncertainty, as only a part of the stream can be stored. In this paper, we evaluate a statistical technique which biases the estimation of the support of patterns, so as to maximize either the precision or the recall, as chosen by the user, a...

Many works deal with the subsequence matching problem using automata structures. It is to decide, given two sequences $s$ and $t$, whether $s$ is a subsequence of $t$. Automata like the Directed Acyclic Subsequence Graph (DASG) or the Subsequence Automaton (SA) accept all subsequences of a set of texts. We focus on this last structure and provide s...

Résumé. Récemment la communauté Extraction de Connaissances s'est inté-ressée à de nouveaux modèles où les données arrivent séquentiellement sous la forme d'un flot rapide et continu, i.e. les data streams. L'une des particularités importantes de ces flots est que seule une quantité d'information partielle est disponible au cours du temps. Ainsi ap...

Data streams - dataflows in which the information arrives in a timely manner - have recently become a major subfield of knowledge extraction. One of their most important singularity is that only a part of the information remains available at a time, which makes it necessary to cope with uncertainty. In this paper, we introduce a novel statistical a...

In this paper, we devise a method for the estimation of the true support of itemsets on data streams, with the objective to maximize one chosen criterion among {precision, recall} while ensuring a degradation as reduced as possible for the other criterion. We discuss the strengths, weaknesses and range of applicability of this method that relies on...

A statistical technique is developed for estimating the sup- port of itemsets on data streams, regardless of the size of the data stored. This technique, which is computationally ultra fast, does not depend on the algorithm used to build or maintain the itemsets. On frequent item- sets, it allows to maximize either the precision or the recall, as c...

When we mine for sequential patterns on a whole data stream it's necessary to cope with uncertainty as only a part of the stream is stored. Even if different approaches exist for mining sequential patterns in static databases they are not suitable for this context. We evaluate a statistical technique which biases the estimation of the support of se...

The modelling of a mobile robot environment is mainly motivated first to obtain a map of the free space (in order to compute safe trajectories for the vehicle) and second to provide the control loop with feed-back of obstacle information. Here, the authors only consider the problem of local representations for the displacements of a vehicle with tw...

This paper presents a new method for D scene analysis with a multi-sensor system. The basic idea is to limit the search of 3D features to interesting regions selected in grey level picture of the scene. Information on the relief of these regions is obtained by using a laser projector; a small number of reflected slits are sufficient to identify and...