Multidimensional Database

D'après une enquête de l'IDC auprès de 45 organisations ayant un Data Warehouse en fonctionnement (fin 1995-1996) : » 90% des entreprises ont un RSI au moins égal à 40% » 50% ont un RSI supérieur à 160% » 25% ont un RSI supérieur à 600% 232323 <#> 2. DW-Stockage Structure de cumul simple

The metric space model abstracts many proximity or similarity problems, where the most frequently considered primitives are range and k-nearest neighbor search, leaving out the similarity join, an extremely important primitive. In fact, despite the great attention that this primitive has received in traditional and even multidimensional databases, little has been done for general metric databases. We solve two variants of the similarity join problem: (1) range joins: Given two sets of objects and a distance threshold r, find all the object pairs (one from each set) at distance at most r; and (2) k-closest pair joins: Find the k closest object pairs (one from each set). For this sake, we devise a new metric index, coined List of Twin Clusters (LTC), which indexes both sets jointly, instead of the natural approach of indexing one or both sets independently. Finally, we show how to use the LTC in order to solve classical range queries. Our results show significant speedups over the basic quadratic-time naive alternative for both join variants, and that the LTC is competitive with the original list of clusters when solving range queries. Furthermore, we show that our technique has a great potential for improvements.

Bitmap indexes must be compressed to reduce input/output costs and minimize CPU usage. To accelerate logical operations (AND, OR, XOR) over bitmaps, we use techniques based on run-length encoding (RLE), such as Word-Aligned Hybrid (WAH) compression. These techniques are sensitive to the order of the rows: a simple lexicographical sort can divide the index size by 9 and make indexes several times faster. We investigate row-reordering heuristics. Simply permuting the columns of the table can increase the sorting efficiency by 40%. Secondary contributions include efficient algorithms to construct and aggregate bitmaps. The effect of word length is also reviewed by constructing 16-bit, 32-bit and 64-bit indexes. Using 64-bit CPUs, we find that 64-bit indexes are slightly faster than 32-bit indexes despite being nearly twice as large.

An Object Oriented Conceptual Model Data Cube is described, on the basis of an example for storage, extraction and analysis of data derived from tests of ship models, conducted in the Bulgarian Ship Hydrodynamics Center -Varna. The main storage parameters as well as the structure of the system are presented. The dynamical level in an object oriented class definition of data cube is described. Future problems of OLAP operations and data analysis methods, included in the UML class definition are outlined.

Partitioned Optimal Passive Stars network, POPS(d, g), is an optical interconnection network of N processors (N = dg) which uses g 2 optical passive star couplers. The processors of this network are partitioned into g groups of d processors each and the g 2 couplers are used for connecting each group with each of the groups, including itself. In this paper, we present an optimal embedding of the hypercube on this network for all combinations of values of d and g. Specifically, we show how to optimally simulate the most common hypercube communication pattern where each hypercube node sends a packet along the same dimension. Optimal simulation of this communication on the POPS(d, g) network has already been presented for d ≤ g in the literature, but for the case d > g, the optimality remained an open problem. Now, we show that an optimal simulation is feasible in this case too.

NoSQL (Not Only SQL) systems are becoming popular due to known advantages such as horizontal scalability and elasticity. In this paper, we study the implementation of data warehouses with document-oriented NoSQL systems. We propose mapping rules that transform the multidimensional data model to logical document-oriented models. We consider three different logical models and we use them to instantiate data warehouses. We focus on data loading, model-to-model conversion and OLAP cuboid computation.

NoSQL (Not Only SQL) systems are becoming popular due to known advantages such as horizontal scalability and elasticity. In this paper, we study the implementation of multidimensional data warehouses with columnoriented NoSQL systems. We define mapping rules that transform the conceptual multidimensional data model to logical column-oriented models. We consider three different logical models and we use them to instantiate data warehouses. We focus on data loading, model-to-model conversion and OLAP cuboid computation.

Dealing with large amount of data has always been a key focus of the Multidimensional Database (MDB) community, especially in the current era when data volume increases more and more rapidly. In this paper, we outline a conceptual modeling solution allowing reducing data in MDBs. A MDB after reduction is modeled with multiple states. Each state is valid during a period of time and aggregates data from a more recent state. We propose three alternatives of reduced MDB modeling at the logical level: (i) the flat modeling integrates all states into one single table, (ii) the horizontal modeling converts each state into a fact table and some dimension tables associated with a temporal interval and (iii) the vertical modeling breaks down a reduced MDB into separate tables, each table includes data from one or several states. We evaluate query execution efficiency in MDBs with and without data reduction. The result shows data reduction is an interesting solution, since it significantly decreases execution costs by 98.96% during our experimental assessments.

Actuellement, si l'intuition reste fondamentale dans le processus de prise de décision, elle ne suffit plus, il faut aussi pouvoir prendre ce que les anglophones appellent « la décision informée » (informed décision). A cette fin, les entreprises disposent aujourd'hui d'importants volumes d'informations qui sont généralement hétérogènes : données structurées (bases de données, fichiers…), données semi-structurées et multimédia (XML, HTML…). Vu cette hétérogénéité des données, il est nécessaire de les contrôler, les homogénéiser, les organiser, les intégrer et enfin les stocker pour donner à leur utilisateur une vue intégrée et orientée métier ; ces informations factuelles et aussi textuelles constituent un facteur de savoir faire et de connaissances. Ainsi, face à l'augmentation considérable du nombre de documents gérés par les entreprises, le besoin d'outils pour traiter automatiquement les informations documentaires s'est rapidement fait sentir par les entreprises. Dans ce contexte, un nouveau concept est apparu permettant de gérer cette masse importante d'informations, à savoir : l'entrepôt de documents. C'est « une source d'informations orientées-sujets, filtrées, intégrées, historisées et organisées comme support d'un processus de recherche, d'interrogation ou d'analyse. » [7]. D'après cette définition, les données d'un entrepôt doivent être organisées par sujets, jugées pertinentes, intégrées en provenance de multiples sources et historisées, c'est à dire garder leur évolution dans le temps, ainsi que leurs différentes versions ; ce dernier point fera l'objet de ce papier. Ce papier se décompose comme suit. La section 2 décrit les travaux existants pour la gestion des versions des documents. La section 3 présente le contexte de nos travaux. Plus précisément, elle décrit le modèle générique proposé d'entrepôts de documents. Dans la section suivante, nous présentons l'extension que nous proposons à ce modèle pour tenir compte de la gestion des versions de documents. Enfin, nous décrivons les modifications apportées à notre outil DOCWARE (DOCument WAREhouse). 2 Etat de l'art Un des problèmes liés à Internet est qu'il permet à l'information de circuler et d'évoluer de manière totalement aléatoire. Ces changements rapides et à priori imprédictibles posent donc le problème de leur détection : un utilisateur visitant des documents de manière répétée peut vouloir être informé de la manière dont le document a été modifié depuis sa dernière visite. Il s'agit donc de pouvoir détecter les changements dus à ces évolutions, en comparant les anciennes et les nouvelles versions du document, et en évaluant leur similarité. Dans ce contexte, plusieurs outils et travaux ont été proposés dans la littérature tels que : XML TreeDiff [4], XyDiff [3], X-Diff [15], [10] et [12]. XML TreeDiff [4] réalisé par IBM, est un ensemble de modules permettant la différentiation et la mise à jour des arbres DOM [14], similaires aux outils diff et patch de différentiation et de mise à jour des fichiers de données classiques. Cependant, au lieu de différencier les représentations des fichiers des documents XML, XML TreeDiff utilise leurs DOM. XyDiff [3] est un composant de Xylème [1] permettant de gérer les différentes versions d'un même document. Tout élément modifié est alors représenté comme un fichier XML, indexé et stocké dans un entrepôt de données. Ces fichiers sont ensuite utilisés pour reconstruire une version antérieure d'un document. XyDiff utilise la structure arborescente des documents XML afin de détecter les mouvements et les changements qui interviennent sur un document.

Many models have been proposed for modeling multidimensional data warehouses and most consider a same function to determine how measure values are aggregated according to different data detail levels. We provide a conceptual model that supports (1) multiple aggregations, associating to the same measure a different aggregation function according to analysis axes or hierarchies, and (2) differentiated aggregation, allowing specific aggregations at each detail level. Our model is based on a graphical formalism that allows controlling the validity of aggregation functions (distributive, algebraic or holistic). We also show how conceptual modeling can be used, in an R-OLAP environment, for building lattices of pre-computed aggregates.

The catalogue metaphor is in widespread use for information systems which divide large document collections into subcategories to support browsing and searching [Marchionini 95]in these collections. In most catalogue applications the document categories are structured in a hierarchical way. The semantic organization of the catalogue's hierarchies plays a crucial role for the navigation efficiency and also for the effectiveness of the catalogue-oriented browsing user interface as a document exploration tool.

One of the important issues in data warehousing is the selection of a set of views to materialize in order to minimize the cost. Materialized view can provide the massive improvement in query processing and it is the crucial decision in a database for optimal efficiency. Materialized views have been found to be very effective at speeding-up queries and are increasingly being supported by commercial databases or data warehouse systems. However, one encounters the problem of view maintenance if all possible views are ...

In this article we test and evaluate the implementation of a dimension of spatio-temporal topological operators within a hypercube, a multidimensional database (MDDB) structure formed by the conjunction of several dimensions. Our goal is to permit adequate Spatio-Temporal Exploration and Analysis (STEA) in the context of Automatic Position Reporting System (APRS), the amateur radio system for position report transmission. Mobile APRS stations are equipped with GPS navigation systems to provide real-time positioning reports. The multidimensional approach has already demonstrated a good potential for spatiotemporal exploration and analysis despite its lack of explicit spatial, temporal and spatio-temporal topological operators that are of key importance for spatio-temporal knowledge discovery. Our approach implements spatio-temporal topological operators in multidimensional databases (MDDBs) through a hierarchy of topological operators representing spatial and temporal relationships between instances of objects, this hierarchy covers the three possible domains of spatio-temporal topological constrains e.g. spatial, temporal and spatio-temporal. At the root of the hierarchy users can make use of simple operators such as "same place", "same time" or "same time, same place" in their multidimensional query. As they

This paper presents the development of a virtual health care system. We suggest that the physician treat patients administer programs that assist both the physician and the patient by rescheduling the appointments and monitoring the medical progress. The virtual system helps and allows the doctors to manage their online reservations. Patients can take online appointments. Number of appointments for different date and time is managed by the portal. E-ASHWINA is a communication link between the doctor and the patients; the opportunity to communicate on a one-to-one basis is highly valued. The administrator has the right to register certified doctors and hospitals on the portal, any third party cannot access data of the portal. Our program also contains a donor module. This portal allows quick search of blood banks nearby your locations and also blood donation centers near you. It provides Covid-19 guidelines for the patients and vaccines information, How can register for vaccines through governments official sites.

Most of the relational database management systems have built-in tuning tools that recommend indexes to be created on tables. These tools consider queries of the single database. They do not support queries that are based on tables of multiple databases within the same relational DBMS or tables of multiple heterogeneous relational DBMSs. In this demonstration, the system "Table Index Evaluator and Recommender (TIER)" is presented which analyzes and recommends table indexes for queries that retrieve data from multiple heterogeneous distributed databases. The system takes set of queries as input, parses queries, analyzes and evaluates existing indexes and recommends new set of indexes.

To support their analytical processes, today's organizations deploy data warehouses and client tools such as OLAP (On-Line Analytical Processing) to access, visualize, and analyze their integrated, aggregated and summarized data. Since a large part of these data have a spatial component, better client tools are required to take full advantage of the geometry of the spatial phenomena or objects being analyzed. With this regard, Spatial OLAP (SOLAP) technology offers promising possibilities. A SOLAP tool can be defined as "a type of software that allows rapid and easy navigation within spatial databases and that offers many levels of information granularity, many themes, many epochs and many display modes synchronized or not: maps, tables and diagrams" [

-les systèmes OLAP facilitent l'analyse en offrant un espace multidimensionnelle des données que les décideurs explorent interactivement par une succession d'opérations Olap. Cependant ces systèmes OLAP sont élaborés pour un groupe de décideurs ou un sujet d'analyse « subject-oriented », pour lesquels sont présumés des besoins parfaitement identiques. Ce qui les rend mal adapté à un usage particulier. La personnalisation vise à mieux prendre en compte l'usager ; ce papier présente une récapitulation sur tous les travaux entrepris dans ce sens avec une étude comparative.

L'analyse en ligne (OLAP) dans les cubes de textes nécessite la définition de nouveaux types d'opérateurs d'analyse appropriés aux données textuelles. En effet, les opérateurs d'agrégation classiques ont montré leur efficacité pour l'analyse en ligne des données numériques, mais ils sont inadaptés pour l'analyse des données textuelles. Dans cet article, nous proposons un nouvel opérateur d'agrégation par catégorisation nommé TLabel (Text Label) permettant d'agréger les données textuelles en plusieurs classes de documents. A chaque classe sera associée une étiquette (Label) qui représente le contenu sémantique des données textuelles de la classe grâce à une adaptation des techniques de fouille de textes à l'OLAP. Nous avons effectué une étude expérimentale sur notre opérateur TLabel. Les résultats préliminaires montrent l'intérêt de notre approche pour l'analyse en ligne des données textuelles.

Mining frequent itemsets (FIs) has been developing in recent years. However, little attention has been paid to efficient methods for mining in multidimensional databases. In this paper, we propose a new method with a supporting structure called AIO-tree (Attributes Itemset Object identifications-tree) for mining FIs from multidimensional databases. This method need not transform the database into the transaction database, and it is based on the intersections of object identifications for fast computing the support of itemsets. We compare our method to dEclat (after transformation to a transaction database) and indeed claim that they are faster than dEclat.

Les données d'un entrepôt sont rafraîchies périodiquement et conservées de manière permanente. Cependant, les décideurs portent généralement un intérêt moindre pour les données anciennes. Dans cet article, nous proposons un mécanisme permettant de synthétiser les données les plus anciennes. Nous définissons un modèle conceptuel d'archivage de données multidimensionnelles. Nous présentons, ensuite, le modèle logique correspondant et les principes permettant d'interroger des schémas multidimensionnels archivés.

Cet article définit un modèle à contraintes pour les bases multidimensionnelles. Notre modèle représente les données en une constellation de faits (sujets d'analyse) associés à des dimensions (axes d'analyse) pouvant être partagées. Chaque dimension peut être structurée en plusieurs hiérarchies (perspectives d'analyse) intégrant différents niveaux de granularités de données. Afin d'assurer la cohérence des données et une plus grande fiabilité de leur manipulation, nous intégrons 5 contraintes sémantiques (exclusion, inclusion, partition, simultanéité, totalité) pouvant être intra-dimension ou inter-dimensions ; les contraintes intra-dimension permettent d'exprimer des contraintes entre les hiérarchies d'une même dimension tandis que les contraintes inter-dimensions portent sur des hiérarchies de dimensions distinctes. Nous étudions également les répercussions de ces contraintes sur les manipulations multidimensionnelles et nous proposons d'étendre les opérateurs multidimensionnels. ABSTRACT. This paper defines a constraint-based model dedicated to multidimensional databases. The model we define represents data through a constellation of facts (subjects of analyse) associated to dimensions (axis of analyse), which are possibly shared. Each dimension is organised according to several hierarchies (views of analyse) integrating several levels of data granularity. In order to insure data consistency, we introduce 5 semantic constraints (exclusion, inclusion, partition, simultaneity, totality) which can be intra-dimension or inter-dimensions; the intra-dimension constraints allow the expression of constraints between hierarchies within a same dimension whereas the inter-dimensions constraints focus on hierarchies of distinct dimensions. We also study repercussions of these constraints on multidimensional manipulations and we provide extensions of the multidimensional operators. MOTS-CLÉS : Bases de données multidimensionnelles, contraintes, opérateurs algébriques.

Dealing with large amount of data has always been a key focus of the Multidimensional Database (MDB) community, especially in the current era when data volume increases more and more rapidly. In this paper, we outline a conceptual modeling solution allowing reducing data in MDBs. A MDB after reduction is modeled with multiple states. Each state is valid during a period of time and aggregates data from a more recent state. We propose three alternatives of reduced MDB modeling at the logical level: (i) the flat modeling integrates all states into one single table, (ii) the horizontal modeling converts each state into a fact table and some dimension tables associated with a temporal interval and (iii) the vertical modeling breaks down a reduced MDB into separate tables, each table includes data from one or several states. We evaluate query execution efficiency in MDBs with and without data reduction. The result shows data reduction is an interesting solution, since it significantly decreases execution costs by 98.96% during our experimental assessments.

Many models have been proposed for modeling multidimensional data warehouses and most consider a same function to determine how measure values are aggregated according to different data detail levels. We provide a conceptual model that supports (1) multiple aggregations, associating to the same measure a different aggregation function according to analysis axes or hierarchies, and (2) differentiated aggregation, allowing specific aggregations at each detail level. Our model is based on a graphical formalism that allows controlling the validity of aggregation functions (distributive, algebraic or holistic). We also show how conceptual modeling can be used, in an R-OLAP environment, for building lattices of pre-computed aggregates.

Catastrophic Interference is a well known problem of Artificial Neural Networks (ANN) learning algorithms where the ANN forget useful knowledge while learning from new data. Furthermore the structure of most neural models must be chosen in advance. In this paper we introduce a hybrid algorithm called Flexible Architecture of Self Organizing Maps (FASOM) that overcomes the Catastrophic Interference and preserves the topology of Clustered data in changing environments. The model consists in K receptive fields of self organizing maps. Each Receptive Field projects high-dimensional data of the input space onto a neuron position in a low-dimensional output space grid by dynamically adapting its structure to a specific region of the input space. Furthermore the FASOM model automatically finds the number of maps and prototypes needed to successfully adapt to the data. The model has the capability of both growing its structure when novel clusters appears and gradually forgets when the data volume is reduced in its receptive fields. Finally we show the capabilities of our model with experimental results using synthetic sequential data sets and real world data.

In the recent past, there has been an irmeas~g interest in multidimensional databases (MDB) and On-line Analytical Processing (OLAP) scenarios. Several multidimensional models have been proposed in the last days. However, very few works have been focused on the area of multidimensional database conceptual modeling. Moreover, they are either conceptual extensions to the classical multidimensional model or translations tkom classical database conceptual models (such as the Entity-Relationship model). Nevertheless, we take the concepts and basic ideas of the classical multidimensional model (dimensions and facts) to propose a revolutionary approach based on the Object Oriented (00) Paradigm to MDB conceptual modeling. Then, the basic elements of our Object Oriented Multidimensional Model (OOMD) such as dimension ckz.~~ and fact classes are introduced. We then present cube cZasses as the basic structure to allow a subsequent analysis of the data stored in the system We fairly believe that the utilization of the 00 Paradigm will provide us a general conceptual model to MDB conceptual modeling in a more flexible, natural and simple way than the models proposed until now.

This paper presents the development of a virtual health care system. We suggest that the physician treat patients administer programs that assist both the physician and the patient by rescheduling the appointments and monitoring the medical progress. The virtual system helps and allows the doctors to manage their online reservations. Patients can take online appointments. Number of appointments for different date and time is managed by the portal. E-ASHWINA is a communication link between the doctor and the patients; the opportunity to communicate on a one-to-one basis is highly valued. The administrator has the right to register certified doctors and hospitals on the portal, any third party cannot access data of the portal. Our program also contains a donor module. This portal allows quick search of blood banks nearby your locations and also blood donation centers near you. It provides Covid-19 guidelines for the patients and vaccines information, How can register for vaccines through governments official sites.

Data cube computation is one of the most essential but expensive operations in data warehousing. Previous studies have developed two major approaches, top-down versus bottom-up. The former, represented by the MultiWay Array Cube (called the MultiWay) algorithm [30], aggregates simultaneously on multiple dimensions; however, it cannot take advantage of a priori pruning [2] when computing iceberg cubes (cubes that contain only aggregate cells whose measure values satisfy a threshold, called the iceberg condition). The latter, represented by BUC [6] , computes the iceberg cube bottom-up and facilitates a priori pruning. BUC explores fast sorting and partitioning techniques; however, it does not fully explore multidimensional simultaneous aggregation. In this paper, we present a new method, Star-Cubing, that integrates the strengths of the previous two algorithms and performs aggregations on multiple dimensions simultaneously. It utilizes a star-tree structure, extends the simultaneous aggregation methods, and enables the pruning of the group-bys that do not satisfy the iceberg condition. Our performance study shows that Star-Cubing is highly efficient and outperforms the previous methods. Index Terms-Data warehouse, data mining, online analytical processing (OLAP). Ç 1 INTRODUCTION E FFICIENT computation of data cubes has been one of the focusing points in research since the introduction of data warehousing, OLAP, and data cube [9]. Previous studies can be classified into the following categories:

Multidimensional (MD) modeling is the basis for Data warehouses (DW), multidimensional databases (MDB) and On-Line Analytical Processing (OLAP) applications. In this paper, we present how the Unified Modeling Language (UML) can be successfully used to represent both structural and dynamic properties of these systems at the conceptual level. The structure of the system is specified by means of a UML class diagram that considers the main properties of MD modeling with minimal use of constraints and extensions of the UML. If the system to be modeled is too complex, thereby leading us to a considerable number of classes and relationships, we describe how to use the package grouping mechanism provided by the UML to simplify the final model. Furthermore, we provide a UML-compliant class notation (called cube class) to represent OLAP users' initial requirements. We also describe how we can use the UML state and interaction diagrams to model the behavior of a data warehouse system. To facilitate the interchange of conceptual MD models, we provide a Document Type Definition (DTD) which allows us to represent the same MD modeling properties that can be considered by using our approach. From this DTD, we can directly generate valid eXtensible Markup Language (XML) documents that represent MD models at the conceptual level. We believe that our innovative approach provides a theoretical foundation for simplifying the conceptual design of MD systems and the examples included in this paper clearly illustrate the use of our approach.

Multidimensional analysis requires the computation of many aggregate functions over a large volume of collected data. To provide the various viewpoints for the analysts, these data are organized as a multi-dimensional data model called data cubes. Each cell in a data cube represents a unique set of values for the different dimensions and contains the metrics of interest. The different abstraction and concretization associated with a dimension may be represented as lattice. The focus is to move up and drill down within the lattice using an algorithm with optimal space and computation. In the lattice of cuboids, there exist multiple paths for summarization from a lower to an upper level of cuboid. The alternate paths involve different amounts of storage space and different volume of computations. Thus objective of this paper is to design an algorithm for formal analysis leading towards detection of an optimal path for any two given valid pair of cuboids at different levels. Algorithm is proposed based on branch and bound method for selection of optimal path. Experimental results in the last show that the solution obtained by the algorithm gives the optimal solution in terms of space and time computation.

This paper presents the development of a virtual health care system. We suggest that the physician treat patients administer programs that assist both the physician and the patient by rescheduling the appointments and monitoring the medical progress. The virtual system helps and allows the doctors to manage their online reservations. Patients can take online appointments. Number of appointments for different date and time is managed by the portal. E-ASHWINA is a communication link between the doctor and the patients; the opportunity to communicate on a one-to-one basis is highly valued. The administrator has the right to register certified doctors and hospitals on the portal, any third party cannot access data of the portal. Our program also contains a donor module. This portal allows quick search of blood banks nearby your locations and also blood donation centers near you. It provides Covid-19 guidelines for the patients and vaccines information, How can register for vaccines through governments official sites.

We proposed a novel approach for face recognition to address the challenging task of recognition using a fusion of nonlinear dimensional reduction; Locally Linear Embedding (LLE) and Principal Component Analysis (PCA) .LLE computes a compact representation of high dimensional data combining the major advantages of linear methods, With the advantages of nonlinear approaches which is flexible to learn a broad of class on nonlinear manifolds. The application of LLE, however, is limited due to its lack of a parametric mapping between the observation and the low-dimensional output. In addition, the revealed underlying manifold can only be observed subjectively. To overcome these limitations, we propose our method for recognition by fusion of LLE and Principal Component Analysis (FLLEPCA) and validate their efficiency. Experiments on CMU AMP Face EXpression Database and JAFFE databases show the advantages of our proposed novel approach.

A multidimensional database stores data as groups of field category values into dimensions, and then groups these dimensions into multidimensional arrays. Specific field category values that may occur in data identify either the rows or columns of array dimensions. The specific grouped field categories themselves identify the row or column array dimensions. This view, when presented to the end user, bring in more relevance and business sense for practical decision making than the views presented by standard relational tables and flat file databases. On-Line Analytical Processing (OLAP) is based on the multidimensional representation of data which is supported by multidimensional databases, relational engines or a combination of both. In this paper we propose a model for multidimensional databases that we had used and implemented.

In multidimensional data models intended for online analytic processing (OLAP), data are viewed as points in a multidimensional space. Each dimension has structure, described by a directed graph of categories, a set of members for each category, and a child/parent relation between members. An important application of this structure is to use it to infer summarizability, that is, whether an aggregate view defined for some category can be correctly derived from a set of precomputed views defined for other categories. A dimension is called structurally heterogeneous if two members in a given category are allowed to have ancestors in different categories. In this article, we propose a class of integrity constraints, dimension constraints , that allow us to reason about summarizability in heterogeneous dimensions. We introduce the notion of frozen dimensions which are minimal homogeneous dimension instances representing the different structures that are implicitly combined in a heterogen...

In multidimensional data models intended for online analytic processing (OLAP), data are viewed as points in a multidimensional space. Each dimension has structure, described by a directed graph of categories, a set of members for each category, and a child/parent relation between members. An important application of this structure is to use it to infer summarizability, that is, whether an aggregate view defined for some category can be correctly derived from a set of precomputed views defined for other categories. A dimension is called structurally heterogeneous if two members in a given category are allowed to have ancestors in different categories. In this article, we propose a class of integrity constraints, dimension constraints , that allow us to reason about summarizability in heterogeneous dimensions. We introduce the notion of frozen dimensions which are minimal homogeneous dimension instances representing the different structures that are implicitly combined in a heterogen...

Data warehouses, multidimensional databases, and OLAP tools are based on the multidimensional (MD) modeling. Lately, several approaches have been proposed to easily capture main MD properties at the conceptual level. These conceptual MD models, together with a precise management of metadata, are the core of any related tool implementation. However, the broad diversity of MD models and management of metadata justifies the necessity of a universally understood standard definition for metadata, thereby allowing different tools to share information in an easy form. In this paper, we make use of the Common Warehouse Metamodel (CWM) to represent the main MD properties at the conceptual level in terms of CWM metadata. Then, CWM-compliant tools could interoperate by exchanging their CWM-based metadata in a commonly understood format and benefit of the expressiveness of the MD model at the conceptual level.

Catastrophic Interference is a well known problem of Artificial Neural Networks (ANN) learning algorithms where the ANN forget useful knowledge while learning from new data. Furthermore the structure of most neural models must be chosen in advance. In this paper we introduce a hybrid algorithm called Flexible Architecture of Self Organizing Maps (FASOM) that overcomes the Catastrophic Interference and preserves the topology of Clustered data in changing environments. The model consists in K receptive fields of self organizing maps. Each Receptive Field projects high-dimensional data of the input space onto a neuron position in a low-dimensional output space grid by dynamically adapting its structure to a specific region of the input space. Furthermore the FASOM model automatically finds the number of maps and prototypes needed to successfully adapt to the data. The model has the capability of both growing its structure when novel clusters appears and gradually forgets when the data volume is reduced in its receptive fields. Finally we show the capabilities of our model with experimental results using synthetic sequential data sets and real world data.

Catastrophic Interference is a well known problem of Artificial Neural Networks (ANN) learning algorithms where the ANN forget useful knowledge while learning from new data. Furthermore the structure of most neural models must be chosen in advance. In this paper we introduce a hybrid algorithm called Flexible Architecture of Self Organizing Maps (FASOM) that overcomes the Catastrophic Interference and preserves the topology of Clustered data in changing environments. The model consists in K receptive fields of self organizing maps. Each Receptive Field projects high-dimensional data of the input space onto a neuron position in a low-dimensional output space grid by dynamically adapting its structure to a specific region of the input space. Furthermore the FASOM model automatically finds the number of maps and prototypes needed to successfully adapt to the data. The model has the capability of both growing its structure when novel clusters appears and gradually forgets when the data volume is reduced in its receptive fields. Finally we show the capabilities of our model with experimental results using synthetic sequential data sets and real world data.

We present a bioinspired algorithm which performs dimensionality reduction on datasets for visual exploration, under the assumption that they have a clustered structure. We formulate a decision-making strategy based on foraging theory, where a software agent is viewed as an animal, a discrete space as the foraging landscape, and objects representing points from the dataset as nutrients or prey items. We apply this algorithm to artificial and real databases, and show how a multi-agent system addresses the problem of mapping highdimensional data into a two-dimensional space.

OLAP users heavily rely on visualization of query answers for their interactive analysis of massive amounts of data. Very often, these answers cannot be visualized entirely and the user has to navigate through them to find relevant facts. In this paper, we propose a framework for personalizing OLAP queries. In this framework, the user is asked to give his (her) preferences and a visualization constraint, that can be for instance the limitations imposed by the device used to display the answer to a query. Given this, for each query, our method computes the part of the answer that respects both the user preferences and the visualization constraint. In addition, a personalized structure for the visualization is proposed.

This paper is interested in the graphical manipulation of data mart schemes described in XML and issued from a generation module of multidimensional models. This manipulation is performed through a set of operations we have defined. These operations consist in adding, deleting and renaming the multidimensional elements.

Data warehouses, multidimensional databases, and OLAP tools are based on the multidimensional (MD) modeling. Lately, several approaches have been proposed to easily capture main MD properties at the conceptual level. These conceptual MD models, together with a precise management of metadata, are the core of any related tool implementation. However, the broad diversity of MD models and management of metadata justifies the necessity of a universally understood standard definition for metadata, thereby allowing different tools to share information in an easy form. In this paper, we make use of the Common Warehouse Metamodel (CWM) to represent the main MD properties at the conceptual level in terms of CWM metadata. Then, CWM-compliant tools could interoperate by exchanging their CWM-based metadata in a commonly understood format and benefit of the expressiveness of the MD model at the conceptual level.

Most conventional learning algorithms require both positive and negative training data for achieving accurate classification results. However, the problem of learning classifiers from only positive data arises in many applications where negative data are too costly, difficult to obtain, or not available at all. This paper describes a new machine learning approach, called ILoNDF (Incremental data-driven Learning of Novelty Detector Filter). The approach is inspired by novelty detection theory and its learning method, which typically requires only examples from one class to learn a model. One advantage of ILoNDF is the ability of its generative learning to capture the intrinsic characteristics of the training data by continuously integrating the information relating to the relative frequencies of the features of training data and their co-occurrence dependencies. This makes ILoNDF rather stable and less sensitive to noisy features which may be present in the representation of the positive data. In addition, ILoNDF does not require extensive computational resources since it operates on-line without repeated training, and no parameters need to be tuned. In this study we mainly focus on the robustness of ILoNDF in dealing with high-dimensional noisy data and we investigate the variation of its performance depending on the amount of data available for training. To make our study comparable to previous studies, we investigate four common methods: PCA residuals, Hotelling's T 2 test, an auto-associative neural network, and a one-class version of the SVM classifier (lately a favored method for one-class classification). Experiments are conducted on two real-world text corpora: Reuters and WebKB. Results show that ILoNDF tends to be more robust, is less affected by initial settings, and consistently outperforms the other methods.

Most of the recent work on adaptive processing and continuous querying of data streams assume that data objects come in the form of tuples, thus relying on the relational data model and traditional relational operators as basis for query processing techniques. Complex types of objects, such as multidimensional data sets or the vast amounts of raster image data continuously streaming down to Earth from satellites have not been considered. In this paper, we introduce a data and query model as a comprehensive and practically relevant basis for managing and querying streams of remotely-sensed geospatial image data. Borrowing basic concepts from Image Algebra, we detail a data model that reflects basic properties of such streams of imagery. We present a query model that includes stream restrictions, transforms, and compositions, and provides a sound basis for formulating expressive and practically relevant queries over streams of image data. Finally, we outline how the data and query model is currently realized in a data stream management system for geospatial image data that supports geographic applications.

We present a bioinspired algorithm which performs dimensionality reduction on datasets for visual exploration, under the assumption that they have a clustered structure. We formulate a decision-making strategy based on foraging theory, where a software agent is viewed as an animal, a discrete space as the foraging landscape, and objects representing points from the dataset as nutrients or prey items. We apply this algorithm to artificial and real databases, and show how a multi-agent system addresses the problem of mapping highdimensional data into a two-dimensional space.

It is commonly agreed that multidimensional data cubes form the basic logical data model for OLAP applications. Still, there seems to be no agreement on a common model for cubes. In this paper we propose a logical model for cubes based on the key observation that a cube is not a self-existing entity, but rather a view over an underlying data set. We accompany our model with syntactic characterisations for the problem of cube usability. To this end, we have developed algorithms to check whether (a) the marginal conditions of two cubes are appropriate for a rewriting, in the presence of aggregation hierarchies and (b) an implication exists between two selection conditions that involve different levels of aggregation of the same dimension hierarchy. Finally, we present a rewriting algorithm for the cube usability problem.