Bartosz Bębel

Metadata management in a multiversion data warehouse

A data warehouse (DW) is supplied with data that come from external data sources (EDSs) that are production systems. EDSs, which are usually autonomous, often change not only their contents but also their structures. The evolution of external data sources has to be reflected in a DW that uses the sources. Traditional DW systems offer a limited support for handling dynamics in their structures and contents. A promising approach to this problem is based on a multiversion data warehouse (MVDW). In such a DW, every DW version includes a schema version and data consistent with its schema version. A DW version may represent a real state at certain period of time, after the evolution of EDSs or changed user requirements or the evolution of the real world. A DW version may also represent a given business scenario that is created for simulation purposes. In order to appropriately synchronize a MVDW content and structure with EDSs as well as to analyze multiversion data, a MVDW has to manage metadata. Metadata describing a MVDW are much more complex than in traditional DWs. In our approach and prototype MVDW system, a metaschema provides data structures that support: (1) monitoring EDSs with respect to content and structural changes, (2) automatic generation of processes monitoring EDSs, (3) applying the discovered EDS changes to a selected, DW version, (4) describing the structure of every DW version, (5) querying multiple DW versions of interest at the same time, (6) presenting and comparing multiversion query results.

OLAP-Like analysis of time point-based sequential data

Nowadays business intelligence technologies allow to analyze mainly set oriented data, without considering order dependencies between data. Few approaches to analyzing data of sequential order have been proposed so far. Nonetheless, for storing and manipulating sequential data the approaches use either the relational data model or its extensions. We argue that in order to be able to fully support the analysis of sequential data, a dedicated new data model is needed. In this paper, we propose a formal model for time point-based sequential data with operations that allow to construct sequences of events, organize them in an OLAP-like manner, and analyze them. To the best of our knowledge, this is the first formal model and query language for this class of data.

Managing evolution of data warehouses by means of nested transactions

In this paper we address problems of managing data warehouses (DWs) that evolve in time and we demonstrate that transactional maintenance of evolving DWs is inevitable. To this end, we propose a nested transaction model. In this model we define 5 types of transactions each of which is responsible for certain tasks. The tasks and properties of these transactions are characterized in the paper.

Sequential Data Analytics by Means of Seq-SQL Language

Ubiquitous devices and applications generate data, whose natural feature is order. Most of the commercial software and research prototypes for data analytics allow to analyze set oriented data, neglecting their order. However, by analyzing both data and their order dependencies, one can discover new business knowledge. Few solutions in this field have been proposed so far, and all of them lack a comprehensive approach to organize and process such data in a data warehouse-like manner. In this paper, we contribute an SQL-like query language for analyzing sequential data in an OLAP-like manner, its prototype implementation and performance evaluation.

FOCUS: an index for continuous subsequence pattern queries

Recent appearance of the a type of OLAP analysis, the sequential OLAP (or SOLAP) has caused the need for new index structures which support new types of analytical queries. An integral part of processing SOLAP queries is finding sequences which match a user-specified pattern. We call such queries subsequence pattern queries. The contribution of this paper is threefold: first, we propose logical and physical index structure which supports subsequence pattern queries, second, we extend this structure to support aggregation queries and third, we perform performance experiments which show that our solutions offer orders of magnitude improvement over previous state of the art solutions.

Hierarchical materialisation of methods in object-oriented views: design, maintenance, and experimental evaluation

The application of materialised object-oriented views in object-relational data warehousing systems is promising. In this paper we propose a novel technique for the materialisation of method results in object-oriented views, called hierarchical materialisation. When an object used to materialise the result of method m is updated, then m has to be recomputed. This recomputation can use unaffected intermediate materialised results of methods called from m, thus reducing a recomputation time. The hierarchical materialisation technique was implemented and evaluated by a number of experiments concerning methods without input arguments as well as methods with input arguments. The results showed that hierarchical materialisation reduces method recomputation time. Moreover, materialising methods with input arguments of narrow discrete domains introduces only a small time overhead.

Comparison of Genetic and Tabu Search Algorithms in Multiquery Optimization in Advanced Database Systems

In several database applications sets of related queries are submitted together to be processed as a single unit. In all these cases the queries usually have some degree of overlap, i.e. may have common subqueries. Therefore a significant performance improvement can be obtained by optimizing and executing the entire group of queries as a whole, thus avoiding to duplicate the optimization and processing effort for common parts. This has suggested an approach, termed multiquery optimization (MQO) that has been proposed and studied by several authors. In this paper we suggest a new approach to multiple-query optimization based on Genetic and Tabu Search algorithms that ensure the tractability of the problem even for very large size of the queries. To analyze the performance of the algorithms, we have run a set of experiments that allow to understand how the different approaches are sensitive to the main workload parameters.

Searching for Patterns in Sequential Data: Functionality and Performance Assessment of Commercial and Open-Source Systems

Ubiquitous devices and applications generate data that are naturally ordered by time. Thus elementary data items can form sequences. The most popular way of analyzing sequences is searching for patterns. To this end, sequential pattern discovery techniques were proposed in some research contributions and implemented in a few database systems, e.g., Oracle Database, Teradata Aster, Apache Hive. The goal of this work is to assess the functionality of the systems and to evaluate their performance with respect to pattern queries.