Esteban Zimanyi

Spatio-temporal conceptual models: data structures + space + time

Nowadays, many applications need data modeling facilities for the description of complex objects with spatial and/or temporal facilities. Responses to such requirements may be found in Geographic Information Systems (GIS), in some DBMS, or in the research literature. However, most of existing models cover only partly the requirements (they address either spatial or temporal modeling), and most are at the logical level, hence not well suited for database design. This paper proposes a spatio-temporal modeling approach at the conceptual level, called MADS. The proposal stems from the identification of the criteria to be met for a conceptual model. It is advocated that orthogonality is the key issue for achieving a powerful and intuitive conceptual model. Thus, the proposal focuses on highlighting similarities in the modeling of space and time, which enhance readability and understandability of the model.

Hierarchies in a multidimensional model: from conceptual modeling to logical representation

Hierarchies are used in data warehouses (DWs) and on-line analytical processing (OLAP) systems to see data at different levels of detail. However, many kinds of hierarchies arising in real-world situations are not addressed by current OLAP systems. Further, there is still no agreement on a conceptual model for DW and OLAP design that offers both a graphical representation and a formal definition.In this paper, we formally define the MultiDimER model, a conceptual multidimensional model that allows to represent facts with measures as well as the different kinds of hierarchies already classified in our previous work [E. Malinowski, E. Zimanyi, OLAP hierarchies: a conceptual perspective, in: Proceedings of the 16th International Conference on Advanced Information Systems Engineering, 2004, pp. 477-491]. We also present the mapping of such hierarchies to the relational model, as well as their implementation in commercial DW products.

Conceptual Modeling for Traditional and Spatio-Temporal Applications: The MADS Approach

Reference LBD-BOOK-2007-001 URL: http://lbd.epfl.ch/e/MADSflyer.pdf Record created on 2007-06-06, modified on 2016-08-08

Query evaluation in probabilistic relational databases

Abstract This paper describes a generalization of the relational model in order to capture and manipulate a type of probabilistic information. Probabilistic databases are formalized by means of logic theories based on a probabilistic first-order language proposed by Halpern. A sound and complete method is described for evaluating queries in probabilistic theories. The generalization proposed can be incorporated into existing relational systems with the addition of a component for manipulating propositional formulas.

The MurMur project: modeling and querying multi-representation spatio-temporal databases

Reference LBD-ARTICLE-2005-001 URL: http://lbdsun.epfl.ch/e/publications_new/articles.pdf/http://www.sciencedirect.com/science/journal/03064379 Record created on 2007-01-16, modified on 2016-08-08

OLAP Hierarchies: A Conceptual Perspective

OLAP (On-Line Analytical Processing) tools support the decision-making process by giving users the possibility to dynamically analyze high volumes of historical data using operations such as roll-up and drill-down. These operations need well-defined hierarchies in order to prepare automatic calculations. However, many kinds of complex hierarchies arising in real-world situations are not addressed by current OLAP implementations. Based on an analysis of real-world applications and scientific works related to multidimensional modeling, this paper presents a conceptual classification of hierarchies and proposes graphical notations for them based on the ER model. A conceptual representation of hierarchies allows the designer to properly represent users’ requirements in multidimensional modeling and offers a common vision of these hierarchies for conceptual modeling and OLAP tools implementers.

Imperfect Information in Relational Databases

Databases model the real world at two levels: the database schema specifies the structure of the database, and the database extension represents a set of facts or events of the real world. The Appropriate Scheme Assumption [10] expresses that this two-level description is a valid view of the world. Such a distinction between schema and extension is often blurred in artificial intelligence.

A model-driven framework for ETL process development

ETL processes are the backbone component of a data warehouse, since they supply the data warehouse with the necessary integrated and reconciled data from heterogeneous and distributed data sources. However, the ETL process development, and particularly its design phase, is still perceived as a time-consuming task. This is mainly due to the fact that ETL processes are typically designed by considering a specific technology from the very beginning of the development process. Thus, it is difficult to share and reuse methodologies and best practices among projects implemented with different technologies. To the best of our knowledge, no attempt has been yet dedicated to harmonize the ETL process development by proposing a common and integrated development strategy. To overcome this drawback, in this paper, a framework for model-driven development of ETL processes is introduced. The benefit of our framework is twofold: (i) using vendor-independent models for a unified design of ETL processes, based on the expressive and well-known standard for modeling business processes, the Business Process Modeling Notation (BPMN), and (ii) automatically transforming these models into the required vendor-specific code to execute the ETL process into a concrete platform.

Defining ETL worfklows using BPMN and BPEL

Decisional systems are crucial for enterprise improvement. They allow the consolidation of heterogeneous data from distributed enterprise data stores into strategic indicators. An essential component of this data consolidation is the Extract, Transform, and Load (ETL) process. In the research literature there has been very few work defining conceptual models for ETL processes. At the same time, there are currently many tools that manage such processes. However, each tool uses its own model, which is not necessarily able to communicate with the models of other tools. In this paper, we propose a platform-independent conceptual model of ETL processes based on the Business Process Model Notation (BPMN) standard. We also show how such a conceptual model can be implemented using Business Process Execution Language (BPEL), a standard executable language for specifying interactions with web services.

Modeling time from a conceptual perspective

Although many temporal models have been proposed in the literature, there is still need for a conceptual model capturing the essential semantics of time-varying information that is free of implementation concerns. This paper first discusses important criteria to be considered when assessing the OconceptualO quality of a temporal model. Then, it presents the main temporal features of MADS, a spatio-temporal conceptual model. The focus is on identifying issues that are either open or controversial, and discussing alternatives, if any. Finally, it is shown how the model may be implemented on top of either TSQL2 or a traditional, non-temporal data model.