Matthias Jarke

Query Optimization in Database Systems

Efficient methods of processing unanticipated queries are a crucial prerequisite for the success of generalized database management systems. A wide variety of approaches to improve the performance of query evaluation algorithms have been proposed: logic-based and semantic transformations, fast implementations of basic operations, and combinatorial or heuristic algorithms for generating alternative access plans and choosing among them. These methods are presented in the framework of a general query evaluation procedure using the relational calculus representation of queries. In addition, nonstandard query optimization issues such as higher level query evaluation, query optimization in distributed databases, and use of database machines are addressed. The focus, however, is on query optimization in centralized database systems.

Mediator: Towards a Negotiation Support System

MEDIATOR is a negotiation support system (NSS) based onevolutionary systems design (ESD) and database-centered implementation.It supports negotiations by consensus seeking through exchange ofinformation and, where consensus is incomplete, by compromise. Thenegotiation problem is shown --graphically or as relational data inmatrix form-- in three spaces as a mapping from control space to goalspace (and through marginal utility functions) to utility space. Withineach of these spaces the negotiation process is characterized byadaptive change, i.e., mappings of group target and feasible sets bywhich these sets are redefined in seeking a solution characterized by asingle-point intersection between them.This concept is being implemented in MEDIATOR, a data-based micro-mainframeNSS intended to support the players and a human mediator inmulti-player decision situations. Each player employs private andshared database views, using his/her own micro-computer decision supportsystem enhanced with a communications manager to interact with themediator DSS. Sharing of views constitutes exchange of informationwhich can lead towards consensus. The human mediator can supportcompromise, as needed, through use of solution concepts and/orconcession-making procedures in the NSS model base. As a concreteexample, we demonstrate the use of the system for group car buyingdecisions.

An optimizing prolog front-end to a relational query system

An optimizing translation mechanism for the dynamic interaction between a logic-based expert system written in PROLOG and a relational database accessible through SQL is presented. The mechanism makes use of an intermediate language that decomposes the optimization problem and makes the proposed approach target-language independent. It can either facilitate expert system - database interaction, e.g., when integrating expert systems into business systems, or augment existing databases with (external) deductive capabilities.

A Field Evaluation of Natural Language for Data Retrieval

Although a large number of natural language database interfaces have been developed, there have been few empirical studies of their practical usefulness. This paper presents the design and results of a field evaluation of a natural language system-NLS-used for data retrieval.

Communications design for Co-oP: a group decision support system

Decision Support Systems (DSSs), computer-based systems intended to assist managers in preparing and analyzing decisions, have been single-user systems for most of the past decade. Only recently has DSS research begun to study the implications of the fact that most complex managerial decisions involve multiple decision makers and analysts. A number of tools for facilitating group decisions have been proposed under the label Group Decision Support Systems (GDSSs). One of the most important functions of a GDSS is to provide problem-oriented services for communication among decision makers. On the basis of an analysis of the communication requirements in various group decision settings, this paper presents an architecture for defining and enforcing dynamic application-level protocols that organize decision group interaction. The architecture has been implemented on a network of personal computers in Co-oP, a GDSS for cooperative group decision making based on interactive, multiple-criteria decision methods.

Designing a Generalized Multiple Criteria Decision Support System

Decision support systems are of many kinds depending on themodels and techniques employed in them. Multiple criteriadecision making techniques constitute an important class of DSSwith unique software requirements. This paper stresses theimportance of interactive MCDM methods since these facilitatelearning through all stages of the decision making process. Wefirst describe some features of Multiple Criteria Decision SupportSystems ( MCDSSs) that distinguish them from classical DSSs. Wethen outline a software architecture for a MCDSS which has threebasic components: a Dialog Manager, an MCDM Model Manager, and aData Manager. We describe the interactions that occur betweenthese three software components in an integrated MCDSS and outlinea design for the Data Manager which is based on a concept oflevels of data abstraction.

Communications Requirements for Group Decision Support Systems

Abstract:Despite the increasing interest in developing group decision support systems (gdss), it remains unclear how communications between participants of a group problem solving process can be designed, implemented, and utilized in a computer-based distributed GDSS. This paper presents a framework for developing a Communications component for the GDSS. It supports conceptualization of a communication system as being composed of four main modules: the Group Norm Monitor, the Group Norm Filter, the Invocation Mechanism, and the individual decision support system (idss)-to-gdss Document Formatter. In reference to the International Standard Organization (ISo) Open System Architecture, the three first modules are integrated in the Application layer, and the last module in the Presentation layer.

Query processing strategies in the PASCAL/R relational database management system

In the database language PASCAL/R, the programming language PASCAL and concepts based on the relational data model are integrated. The paper investigates transformation strategies used in the PASCAL/R system to evaluate queries with existential and universal quantifiers. Intermediate data structures are described using a high-level language tool called a reference to a selected variable. The predicate calculus approach used in PASCAL/R offers new insight into recently proposed query optimization techniques and allows some of them to be extended.

Logic-Based Formula Management Strategies in an Actuarial Consulting System

In many decision support systems, multiple decision methodsand models must be combined for solving a complex problem.Expertise is required for selecting, adapting and coordinatingappropriate models. This paper describes the design andimplementation of a knowledge-based model management systemcalled the Actuarial Consulting System (ACS). The ACS supportsactuaries in making pricing decisions in the domain of lifeinsurance. Actuarial knowledge is organized using a graphformalism called Formula Derivation Network (FDN), representedin Prolog as a hierarchy of predicates. On the user level, aProblem Analyzer converts a problem specification by the userinto a search problem on the stored collection of FDNs. Usingdifferent search strategies, including human expert rules, theSurface Planner generates an efficient solution strategy(sequence of models). At the lowest level, a Plan Executorretrieves or requests model data and issues appropriate functioncalls to a subroutine library.

Access to Specific Declarative Knowledge by Expert Systems: The Impact of Logic Programming

As part of the operation of an Expert System, adeductive component accesses a database of facts to helpsimulate the behavior of a human expert in a particularproblem domain. The nature of this access is examined, andfour access strategies are identified. Features of each ofthese strategies are addressed within the framework of aLogic-based deductive component and the relational model ofdata.