DSS Concepts, Methodologies, and Technologies PDF
Document Details
Uploaded by ClearedCatharsis
Tags
Summary
This document provides an overview of decision support systems (DSS). It discusses various aspects of DSS, including configurations, characteristics, and capabilities. The document also covers different classifications of DSS and concludes with a brief overview of the future developments and components of DSS.
Full Transcript
Decision Support and Business Intelligence Systems Chapter 3: DSS Concepts, Methodologies, and Technologies Learning Objectives Understand possible decision support system (DSS) configurations Understand the key differences and similarities between DSS and BI systems Describe DSS characteristics an...
Decision Support and Business Intelligence Systems Chapter 3: DSS Concepts, Methodologies, and Technologies Learning Objectives Understand possible decision support system (DSS) configurations Understand the key differences and similarities between DSS and BI systems Describe DSS characteristics and capabilities Understand important DSS classifications Understand DSS components and how they integrate DSS Configurations Many configurations exist; based on ◦ management-decision situation ◦ specific technologies used for support DSS have three basic components 1. Data 2. Model 3. User interface 4. (+ optional) Knowledge DSS Configurations Each component ◦ has several variations; are typically deployed online ◦ Managed by a commercial of custom software Typical types: ◦ Model-oriented DSS (manipulation of a model, for example, statistical, financial, optimization and/or simulation models) ◦ Data-oriented DSS (manipulation of a time- series of internal company data and sometimes external data) DSS Characteristics and Capabilities DSS is not quite synonymous with BI ◦ DSS are generally built to solve a specific problem and include their own database(s) ◦ BI applications focus on reporting and identifying problems by scanning data stored in data warehouses ◦ Both systems generally include analytical tools (BI called business analytics systems) ◦ Although some may run locally as a spreadsheet, both DSS and BI uses Web DSS Characteristics and Capabilities DSS Characteristics and Capabilities Business analytics implies the use of models and data to improve an organization's performance and/or competitive posture Web analytics implies using business analytics on real-time Web information to assist in decision making; often related to e-Commerce Predictive analytics describes the business analytics method of forecasting problems and opportunities rather than simply reporting them as they occur DSS Classifications AIS SIGDSS Classification (Power 2002) ◦ Communication driven and group DSS ◦ Data-driven DSS ◦ Document-driven DSS ◦ Knowledge-driven DSS, data mining and management ES applications ◦ Model-driven DSS ◦ Compound/hybrid DSS *AIS SIGDSS – Association for Information Systems Special Interest Group DSS DSS Classifications Communication driven and group DSS ◦ use collaboration and communication technologies to support groups (support meetings, supply chain management) Data-driven DSS ◦ primarily involved with data, process data into information and present the information to decision maker; database and data warehouse (DSS developed in Data Mining and OLAP) Document-driven DSS ◦ Rely on knowledge coding, analysis, search and retrieval , DSS of text based; minimal emphasis on model; documents (oral, written, multimedia etc.) DSS Classifications Knowledge-driven DSS, data mining and management ES applications ◦ To address specific decision needs; almost all AI apps. Model-driven DSS ◦ Primarily developed around one or more optimization or simulation models; focus is to optimize one or more objectives Compound/hybrid DSS ◦ Include two or more major DSS categories *AIS SIGDSS – Association for Information Systems Special Interest Group DSS DSS Classifications Other DSS Categories ◦ Institutional (portfolio managements system – investment) and ad-hoc DSS ◦ Personal, group, and organizational support ◦ Individual support system versus group support system (GSS) ◦ Custom-made systems versus ready-made systems DSS Classifications Holsapple and Whinston's Classification (2000) ◦ The text-oriented DSS ◦ The database-oriented DSS. ◦ The spreadsheet-oriented DSS ◦ The solver-oriented DSS ◦ The rule-oriented DSS (include most knowledge-driven DSS, data mining, management, and ES applications) ◦ The compound DSS Group Discussion Each group is required to select one DSS category using the AIS SIGDSS scheme. Present the DSS application to your classmate; preferably the recent DSS you have experience with. Components of DSS Components of DSS Data Management Subsystem ◦ Includes the database that contains the data ◦ Database management system (DBMS) ◦ Can be connected to a data warehouse Model Management Subsystem ◦ Model base management system (MBMS) User Interface Subsystem Knowledgebase Management Subsystem ◦ Organizational knowledge base Overall Capabilities of DSS Easy access to data/models/knowledge Proper management of organizational experiences and knowledge Easy to use, adaptive and flexible GUI Timely, correct, concise, consistent support for decision making Support for all who needs it, where and when he/she needs it - See Table 3.2 for a complete list... DSS Components and Web Impacts Impacts of Web to DSS ◦ Data management via Web servers ◦ Easy access to variety of models, tools ◦ Consistent user interface (browsers) ◦ Deployment to PDAs, cell phones, etc. … DSS impact on Web ◦ Intelligent e-Business/e-Commerce ◦ Better management of Web resources and security, … (see Table 3.3 for more…) DSS Components Data Management Subsystem DSS database DBMS Data directory Query facility Database Management Subsystem Key Data Issues Data quality ◦ “Garbage in/garbage out" (GIGO) Data integration ◦ “Creating a single version of the truth” Scalability Data Security Timeliness Completeness, … 10 Key Ingredients of Data (Information) Quality Management 1. Data quality is a business problem, not only a systems problem 2. Focus on information about customers and suppliers, not just data 3. Focus on all components of data: definition, content, and presentation 4. Implement data/information quality management processes, not just software to handle them 5. Measure data accuracy as well as validity 10 Key Ingredients of Data (Information) Quality Management 6. Measure real costs (not just the percentage) of poor quality data/information 7. Emphasize process improvement/preventive maintenance, not just data cleansing 8. Improve processes (and hence data quality) at the source 9. Educate managers about the impacts of poor data quality and how to improve it 10. Actively transform the culture to one that values data quality DSS Components Model Management Subsystem Model base MBMS Modeling language Model directory Model execution, integration, and command processor DSS Components Model Management Subsystem Model base (= database ?) Model Types ◦ Strategic models ◦ Tactical models ◦ Operational models ◦ Analytic models Model building blocks Modeling tools DSS Components Model Management Subsystem The four (4) functions 1. Model creation, using programming languages, DSS tools and/or subroutines, and other building blocks 2. Generation of new routines and reports 3. Model updating and changing 4. Model data manipulation Model directory Model execution, integration and command DSS Components User Interface (Dialog) Subsystem Interface ◦ Application interface ◦ User Interface Graphical User Interface (GUI) DSS User Interface ◦ Portal ◦ Graphical icons Dashboard ◦ Color coding Interfacing with PDAs, cell phones, etc. DSS Components Knowledgebase Management System Incorporation of intelligence and expertise Knowledge components: ◦ Expert systems, ◦ Knowledge management systems, ◦ Neural networks, ◦ Intelligent agents, ◦ Fuzzy logic, ◦ Case-based reasoning systems, and so on Often used to better manage the other DSS components DSS Components Future/current DSS Developments Hardware enhancements ◦ Smaller, faster, cheaper, … Software/hardware advancements ◦ data warehousing, data mining, OLAP, Web technologies, integration and dissemination technologies (XML, Web services, SOA, grid computing, cloud computing, …) Integration of AI -> smart systems DSS User One faced with a decision that an MSS is designed to support ◦ Manager, decision maker, problem solver, … The users differ greatly from each other ◦ Different organizational positions they occupy; cognitive preferences/abilities; the ways of arriving at a decision (i.e., decision styles) User = Individual versus Group Managers versus Staff Specialists [staff assistants, expert tool users, business (system) analysts, facilitators (in a GSS)] DSS Hardware Typically, MSS run on standard hardware Can be composed of mainframe computers with legacy DBMS, workstations, personal computers, or client/server systems Nowadays, usually implemented as a distributed/integrated, loosely-coupled Web- based systems Can be acquired from ◦ A single vendor ◦ Many vendors (best-of-breed) Summary DSS is designed to support complex managerial problems that other computerized techniques cannot DSS – generally developed to solve specific managerial problems, whereas BI systems typically do reporting The AIS SIGDSS classification: data-driven, model- driven, communications-driven, document-driven, knowledge-driven, hybrid DSS Major components of a DSS: database and its management, model base and its management, and a user-friendly interface. Knowledge-based component is an optional