Business intelligence tools for building the executive information systems

Prof. univ. dr. Ion LUNGU, Asist. univ. drd. Adela BRA, Drd. Anca-Georgiana FODOR, Academy of Economic Studies, Bucharest, Romania
Abstract - The Executive Information Systems (EIS) are designed to improve the quality of strategic level of management in organization through a new type of technology and several techniques for extracting, transforming, processing and presenting data in order to provide strategic information. These technologies are known as Business Intelligence Tools. This paper presents the arhitecture of Executive Information Systems and also the main technologies used for designing and building an EIS. Keywords - Business Intelligence (BI), Executive Information Systems (EIS), Datawarehouse, OLAP (On- Line Analytical Processing), Data Mining.
Level of decision

non-ERP application sources. In the table 1 we represent the main differences between ERP reports and EIS reports. Table 1: A comparison of ERP and EIS systems
CHARACTERISTICS Objectives ERP REPORTS Analyse indicators that measure current and internal activities or daily reports EIS REPORTS Processes optimization, analyse key performance indicators, forecast internal and external data, internal and external focus Strategic/High Executives, strategic level of management Datawarehouse OLAP DataMining Analyse Huge Cube Aggregate Historical/current/p rospective

Operational/Medium Operational level of management Relational databases Datawarehouse Report/Analyse Limited Record Detailed, sumarised, pre-aggregate Current

I. INTRODUCTION

User involved

he purpose of Executive Infromation Systems is to assist high level managers in taking strategic decissions and to provide in real time representative informations, to support activities such as goal setting, planning and forecasting, and also tracking performance. Another objective of these systems is to gather, analyze, and integrate internal and external data into dynamic profiles of key performance indicators (KPI). Based on each executives information needs, EIS can access both historical and real-time data through ad-hoc queries. EIS users can manage and manipulate very large sets of data. In essence, managers at every level can have a customized view that extracts information from disparate sources and summarizes it into meaningful indicators. EIS consists in a set of technology solutions that is based on business intelligence (BI) tools. EIS provide a friendly graphical interface and when this is customised for the individual manager, allow users to access corporate data and complements the executive's personal knowledge and provide quantitative diagnostics to monitor the progress of decisions. In many organizations there are implemented ERP systems for operational and transactional processing for different functional areas such as: financials, inventory, purchase, order management, production. Information from this functional areas within a ERP system is managed by a relational software database such as Oracle Database or Microsoft SQL Server. Operational levels of management requires detailed reports with daily operational activities. But executive levels need informations for strategic and tactical decision that often requires reports of agregated data from ERP and

Data Management

Typical operation Number of records/ transaction Data Orientation Level of detail Age of data

II. EIS ARCHITECTURE AND BI TOOLS EIS systems demand for technology solutions that can extract, analyze, and visualize information from ERP and stand-alone systems in real time and with a friendly and flexible user interface. EIS architecture is common to Decision Support Systems architecture and its structured on three distinct levels: Data Management represented by relational databases, datawarehouses and other type of data sources; Model Management wich is the level of extract, transformation and processing of data; Data Visualization Tools that provide a visual drill-down capacity that can help managers examine data graphically and identify complex interrelationships. EIS systems require at each level to apply a set of different technologies wich can provide the best performance in real time querries. These tools and technologies are part of Business Intelligences tools. Well present in the next section the main technologies that are implemented at each level of EISs architecture.

Electronic copy of this paper is available at: http://ssrn.com/abstract=967652

A. Data Management Level This level consists of the data sourcess integration throught datawarehouses. A datawarehouse collects and organizes data from both internal and external sources and makes it available for the purpose of analysis. A datawarehouse contains both historical and current data and it is optimized for fast query and analisys. Data are organized in another type of schema whitch contain fact tables and dimension tables. A fact table is related with dimension tables and contains measures and formula which enable a much easier way in finding data. Dimension tables are structured on different hierarchical levels of aggregation (e.g. Time dimension can have day, week, month and year as hierarchical levels). Data presented in fact tables derrived from different type of data sources like relational databases and user files. Datawarehouses extract, transform and process data for high-level integration and analysis. Datawarehouses architecture is different for each individual organization, but in generally it consists of three levels: data sources, ETL process and data marts like in figure 1:
Data sources BI Tools Datawarehouse Querries Data Integration and ETL Data Warehous e OLAP

detailed information from a multi-dimensional database. Traditional relational database systems handle this situation by using multiple queries. In many cases, the queries become so complex that even the developer finds them difficult to maintain. OLAP overcomes this barrier by enabling users to analyze multi-dimensional data. OLAP Council provided in 1995 this definition: OnLine Analytical Processing (OLAP) is a category of software technology that enables analysts, managers and executives to gain insight into data through fast, consistent, interactive access to a wide variety of possible views of information that has been transformed from raw data to reflect the real dimensionality of the enterprise as understood by the user. OLAP systems have typically been implemented using two technologies: ROLAP (Relational OLAP), where data is stored in a RDBMS and MOLAP (Multidimensional OLAP) where dedicated multidimensional DBMS is used. There are also version of HOLAP (Hybrid OLAP) and DOLAP (Desktop OLAP) systems. Managers can use an OLAP engine for typical operations like slice and dice data by various dimensions and then drill-down into the source data or roll-up to aggreagate levels. OLAP provide tools for forecasting data and what-if scenarious and analysis. But OLAP can only mark the trends and patterns within the data that was requested. It will not discover hidden relationships or patterns, which requires more powerful tools like data mining. D. Data Mining

Data Mining Reports Data Marts

Files Database External sources

Figure 1: Datawarehouses architecture All data sources can be integrated into a central source datawarehouse from where data are extracted, transformed and loaded throught ETL process into a final storage place wich can be a central datawarehouse or many data marts which are departamental datawarehouses. Although a datawarehouse can make it easier and more efficient to use the EIS, it is not required for an EIS to be deployed. Organizations can extract data directly from their host system database for their analysis and reporting purposes, but in a more difficult way. B. Model Management Level At this level we can find BI tools for extracting and analyzing data such as OLAP systems, Data mining process and statistical tools. C. Online Analytical Processing (OLAP) An OLAP engine is a query generator that provides users with the ability to explore and analyze summary and 2

Data mining tools are especially appropriate for large and complex datasets. Through statistical or modeling techniques, data mining tools make it possible to discover hidden trends or rules that are implicit in a large database. Data mining tools can be apply to data from datawarehouses or relational databases. Data discovered by these tools must be validated and verified and then to become oprerational data that can be used in decision process. Data mining tools can answer the questions like How? and Why? in comparison whith the OLAPs question What if?. We can say that Data mining and OLAP systems are complementary and if we want to integrate both technologies into one we can build an OLAM system. OLAM (on-line analytical data mining) systems are OLAP systems used for data mining, used to discover new information from multidimensionaldata. E. Data Visualization Tools Level This level contains tools for presenting and analyzing data from previous levels. There are many graphical tools for building friendly and flexible presentations like: reports, graphics, charts builders, web pages wich can be integrated into an organizational portal or an ERP system interface such as Oracle E-Business Suite.

Electronic copy of this paper is available at: http://ssrn.com/abstract=967652

TOP-TIER Presentations tools: reports, graphs, charts builders, web.

MIDDLE-TIER Data Mining OLAP WEB Technologies

BOTTOM-TIER Metadata Central datawarehouse Data Marts

Data Targets

deployed. Data quality risks relate primarily to whether or not data has been properly cleansed. Technology obsolescence refers to the failure on the part of the vendor to anticipate new technologies. Large budgets and strategic information are involve in deploying EIS systems this is the reason to establish rigorous criteria for evaluating EIS systems. EIS systems have a powerful impact on strategic decisions quality to reduce the time for making decisions. EIS must have the ability to allow managers to view data in different perspective, to drill-down and roll-up to aggregate levels, to navigate and on-line query data sets in order to discover new factors that affect business process and also to anticipate and forecast chenges inside and outside the organization. EIS improve the quality of management in organization through new type of technology and techniques for extracting, transforming, processing and presenting data in order to provide strategic information. REFERENCES [1] Kaniclides, A., Kimble, C. A Development Framework for Executive Information Systems, Proceedings of GRONICS '95, Groningen, The Netherlands, Ed T LOURENS, February 1995 [2] Han, J., Kamber, M. - Data Minig: Concepts and Techniques, Morgan Kaufmann Publishers, San Francisco, 2001 [3] Inmon, W.H. - Building the Data Warehouse, John Wiley & Sons, New York, 1996 [4] Kimball, Ralph The data warehousing toolkit., John Wiley, 1996. [5] Liang, Leo Yonghong; Miranda, Rowan Dashboards and Scorecards: Executive Information Systems for the Public Sector, Government Finance Review, December 2001 [6] Moss L., Atre S. Business Intelligence Roadmap The complete project lifecycle for decision-support applications, Addison-Wesley, 2004 [7] Wheeler, Frederick P. The Potential for Executive Information Systems to Support the Management of Business Reconfiguration, Journal of Financial Information Systems, 1996 [8] *** Internet resources: http://www.oracle.com http://www.intelligenenterprise.com http://www.olapcouncil.org

Extract/Transform/Load (ETL) Integration

Data sources Financials HR Files

Production

Logistics

External sources

Figure 2: EISs architecture EIS should permit the user interface to accommodate different degrees of technical knowledge. An EIS should be designed to allow managers who are not trained to use query languages and advanced technologies, a fast, easy, and understandable way to navigate into data and identify trends and patterns. In the fig. 2 we propose a complex EIS architecture with three distinct levels: bottom tier for data management level, middle tier for model management level and top tier for user interface level. Fig. 2 illustrates an example of an EIS built on a datawarehouse data are processed by an OLAP engine and presented using tables and graphic tools. User has the abillity to select data sets for queries. III. CONCLUSION

Developing EIS systems involve time, high-costs and human resources, efforts and an EIS must be capable to provide in real time representative informations to the executive management. In deploying EIS there are many risks involved: system design, data quality, and technology obsolescence. System design risks stem from poor conceptualization of an enterprises true business needs before the technology is 3