Im Rahmen der Arbeit wird untersucht, wie durch die konsequente Anwendung der objektorientierten Konzepte auf die Architektur von Management Support Systemen (MSS) die beobachtbaren Integrationsbeschränkungen bislang heterogen modellierter MSS-Funktionalitäten überwunden werden können. Zur Identifikation, Auswahl und Konfiguration der diesbezüglichen Potentiale und Konzepte des objektorientierten Paradigmas und späteren Bewertung der daraus entwickelten objektorientierten MSS-Architektur wird zunächst ein dreistufiger Kriterienkatalog aufgestellt, der sich an den spezifischen integrativen Anforderungen an ein MSS orientiert.Das entwickelte MSS-Konzept, das in Form eines konkreten Smalltalk-basierten, objektorientierten MSS (ooMSS) dargestellt wird, umfaßt die drei Modellbereiche Informationsmodell, Interaktionsmodell und MSS-Modell:Das Informationsmodell ist unterteilt in ein operatives Domänenmodell zur Repräsentation heterogener Informationselemente und -mengen und ein Werkzeugmodell zur Abbildung analytischer Beziehungen und Auswertungsfunktionen auf diesen Informationselementen, u.a. mehrdimensionaler Datenanalysen (OLAP). Zur Umsetzung werden die spezialisierten Klassen Information- und ToolObject eingeführt, die über das Adapter- bzw. InterpreterPattern verbunden werden.Das Interaktionsmodell zur direkten, anwenderindividuellen Konfiguration und Manipulation relevanter Ausschnitte des Informationsmodells (Daten und Funktionen) ist nach dem Morphic-Framework realisiert.Das MSS-Modell dient der Repräsentation von technisch und fachlich möglichen bzw. sinnvollen MSS-Funktionen und -abläufe sowie deren inhaltlichen Ergebnissen. Die Modellierung erfolgt mittels der spezialisierten Klasse InformationAspect auf einer Meta-Ebene des Informationsmodells.Abschließend wird die Anwendung des Gesamtkonzepts in Form von MSS-Unterstützungsszenarien auf den Ebenen Fachanwender, Fortgeschrittener und Entwickler dokumentiert.

This thesis describes OLAP and data mining technologies and methods of their communication with users by using dynamic tables. Key theoretical and technical information is also included. Next part focuses on particular implementation of dynamic tables used in Vema portal solution. Last parts come close to analysis and implementation of the metadata editor which enables the metadata to be effectively designed.

OLAP systems are sought-after tool which are deployed in companies and industry environment. The fundamental task of these systems are support of executive management. This work is dealing with OLAP analysis and its capabilities. It also discussed FLEX and AIR technology, their base principles and functions.

This Master's Thesis deals with the integration of Business Intelligence tools into an information system. There are concepts of BI, data warehouses, the OLAP analysis introduced as well as the knowledge discovery from databases, especially the association rule mining. In the chapters focused on practical part of the thesis, the design and implementation of resultant application are depicted. There are also the applied technologies like i.e. Microsoft SQL Server 2005 described.

根據<CFO Magazine>調查，75%的高階主管在面臨擬定策略時，通常無法獲得即時且完整的資訊，以致於在競爭激烈的產業環境中，失去了主動出擊的機會，也錯失了替公司創造財富的機會。 企業要提升競爭力，除了與企業本身運作的良窳相關外，也與整體的資訊支援能力有高度的關係。企業相關資訊系統的資訊如何作適度的整合，以利相關使用者，因此導入整合管理、決策與資訊科技的商業智慧系統，對內可以提升管理績效，對外以形成競爭優勢，實際化資訊為行動，並將其績效評估以指標式呈現，藉助有效的績效管理來提升服務，降低管理成本，以求競爭力的提升。Kaplan and Norton在平衡計分卡一書上說「如果您不能對他進行衡量，您便無法管理它」。 在傳統資訊應用的環境裡，只要使用者一有資訊的需求，就會要求資訊人員設計程式來列印所需的報表，或查詢所需的資訊，但也因為要更深入瞭解報表的內容，所以需要用更多的報表來作補充說明，如此就陷入了所謂的「傳統報表的惡性循環」裡，而使用者卻感覺永遠少一張報表。 「Get the right information，to the right people，at the right time」是商業智慧系統應用的最佳境界，然而其資料的來源，除了外部資訊外，絕大部份是從企業的歷史資料或由現有的應用系統提供，商業智慧是整合企業所有應用系統的資料，以提供給所有的使用者使用，也就是這些應用系統所產生的資料是商業智慧的基礎，有了這些資料，商業智慧才能發揮功效，兩者是相輔相成，企業如能充分發揮商業智慧和應用系統的各自特點，結合應用，必能提高企業的競爭力。 因此本論文提供一個銷售分析的實際案例，以Top Down的思考模式來定義各階層使用者的需求，而實際開發的步驟是以Bottom Up的方式來建置，以整體規劃，分階段執行的建置方式來確保商業智慧系統專案的成功。

Nesta dissertação são feitas três abordagens. Na primeira apresentam-se os componentes necessários para que se possa compreender melhor o cenário atual onde se encontram aqueles que são os responsáveis pelo processo decisório nas empresas. São abordados as evoluções do processo decisório e do suporte ao mesmo, suas etapas e os seus fatores de influência. A segunda abordagem é relativa às três ferramentas que constituem a nova geração de Sistemas de Apoio à Decisão. A primeira ferramenta é o data warehouse, um banco de dados específico para propósitos gerenciais e que é independente dos bancos de dados operacionais. A segunda ferramenta é o On-Line Analytical Processing (OLAP) e é utilizada para se realizar análises sofisticadas, que permitem aos seus usuários compreenderem melhor os negócios que são realizados na empresa. A última ferramenta é o data mining que permite que se faça uma análise nos dados armazenados, durante anos, para que se descubram relacionamentos ocultos entres os dados, revelando perfis de compras e de clientes; desta forma, as informações obtidas podem se tornar estratégias de negócios. Com a abordagem destas três novas ferramentas, deseja-se analisar o que existe de mais avançado, atualmente, para dar um melhor suporte ao processo decisório, sem entrar nos detalhes estritamente técnicos destas tecnologias. A terceira abordagem é constituída de exemplos de empresas que implementaram estas ferramentas e os resultados obtidos, assim como pelas tendências destas ferramentas para os próximos anos. / In this dissertation we will deal with three approaches. On the first we present the necessary elements to make one understand better the current scenery where the responsible persons for the decision process of companies meet. The evolution of the decision process and its support, phases and influence factors. The second approach is related to the three tools that constitute the new generation of Decision Support Systems. The first tool is the data warehouse, a specific database for the managerial purposes that is independent from the operational databases. The second tool is the On-Line Analytical Processing (OLAP) used in carrying out sophisticated analyses allowing its users a better understanding of the business accomplished in the company. The last tool is the data mining that allows for an analysis of the data stored along the years so that one is able to find out the correct relationship among the collect’s data, revealing business and clients profiles. In such way all the information gathered in the process can be converted into business strategy. With the approach of these three new tools we intend to analyze the most advanced techniques available nowadays to give a better decision support without getting into strictly technical details of these technologies. The third approach is made up of examples of companies that implemented such tools and the attained results, as well, the trends for these tools in the coming years.

Business intelligence is an effort to better understanding of company processes and business consequences in which company occur. The main goal of this thesis is to characterize the principles and possibilities of implementation and application Business intelligence in small or medium enterprise. This goal should be achieved by description of general needs which small or medium enterprise, looking forward to use business intelligence, has. This includes evaluation and universal description of information technologies in small and medium enterprises, as well as what data are used for managerial decisions and how are they made. In order to implement BI into the small or medium enterprise, it is necessary to imply how the Business intelligence work and on what principles. The application part of this thesis describes real procedures in preparation of BI implementation, impact on company and evaluation of the meaningfulness of this solution.

Diploma thesis focuses on topic of business performance management realized through Business Intelligence application created on the basis of a balanced set of indicators within the Balanced Scorecard method. The first part introduces the reader with theoretical foundations. First, it describes the basic principles of the BSC method and meaning of dividing strategic objectives into four perspectives. Subsequently, it introduces the heart of BI - a multidimensional approach, displays the BI architecture and defines its components. At the end it compares the types of database systems. There is also a comparison of this thesis with other projects dealing with similar themes in the theoretical part. After the theoretical chapters comes a practical part which aims at design and development of BI applications for a real company so as to comply with the requirements of top managers and the corporate strategy. At the beginning it introduces a firm to which the project is applied. This is an existing organization operating in the area of car sales and services. The corporation is called under a fictitious name "AUTO a.s." in this thesis. In cooperation with its managers the SWOT analysis and strategic objectives is defined. This determines the key requirements for setting up the BI application. An example of a BI solution that allows managers to effectively monitor the development objectives, brings the main benefit of the whole work. For this purpose is built data warehouse, which is then filled with transformed data from source systems using Data Pump. Then, above the Data Warehouse, two OLAP cubes can be created. Presentation layer, created in MS Excel 2007, handles the interpretation of the results to end-users.

Les entrepôts de données et les systèmes d'analyse en ligne OLAP (On-Line Analytical Processing) fournissent des méthodes et des outils permettant l'analyse de données issues des systèmes d'information des entreprises. Mais, seules 20% des données d'un système d'information est constitué de données analysables par les systèmes OLAP actuels. Les 80% restant, constitués de documents, restent hors de portée de ces systèmes faute d'outils ou de méthodes adaptés. Pour répondre à cette problématique nous proposons un modèle conceptuel multidimensionnel pour représenter les concepts d'analyse. Ce modèle repose sur un unique concept, modélisant à la fois les sujets et les axes d'une analyse. Nous y associons une fonction pour agréger des données textuelles afin d'obtenir une vision synthétique des informations issues de documents. Cette fonction résume un ensemble de mots-clefs par un ensemble plus petit et plus général. Nous introduisons un noyau d'opérations élémentaires permettant la spécification d'analyses multidimensionnelles à partir des concepts du modèle ainsi que leur manipulation pour affiner une analyse. Nous proposons également une démarche pour l'intégration des données issues de documents, qui décrit les phases pour concevoir le schéma conceptuel multidimensionnel, l'analyse des sources de données ainsi que le processus d'alimentation. Enfin, pour valider notre proposition, nous présentons un prototype.

L'information géographique est maintenant un produit de masse fréquemment manipulé par des utilisateurs non-experts en géomatique qui ont peu ou pas de connaissances de la qualité des données qu'ils utilisent. Ce contexte accroît significativement les risques de mauvaise utilisation des données et ainsi les risques de conséquence néfaste résultant de ces mauvaises utilisations. Cette thèse vise à fournir à des utilisateurs experts ou des experts en qualité une approche leur permettant d'évaluer la qualité des données et ainsi être à même de conseiller des utilisateurs non-experts dans leur utilisation des données. Cette approche se base sur une structuration des données de qualité dans une base de données multidimensionnelle et une communication dynamique et contextuelle utilisant des indicateurs de qualité affichés dans un système SOLAP (Spatial On-Line Analytical Processing) combiné à un système d'information géographique.