Fred J. Maryanski

Semantic data models

Semantic data models have emerged from a requirement for more expressive conceptual data models. Current generation data models lack direct support for relationships, data abstraction, inheritance, constraints, unstructured objects, and the dynamic properties of an application. Although the need for data models with richer semantics is widely recognized, no single approach has won general acceptance. This paper describes the generic properties of semantic data models and presents a representative selection of models that have been proposed since the mid-1970s. In addition to explaining the features of the individual models, guidelines are offered for the comparison of models. The paper concludes with a discussion of future directions in the area of conceptual data modeling.

Using a metamodel to represent object-oriented data models

A metamodel for describing the semantics of object-oriented data models is presented. This metamodel supports the automatic generation of conceptual database design tool software for a variety of data models. Data modeling is viewed as having three levels of abstraction. The physical database, at the lowest level, stores objects and relationships. Specific data models reside at the middle level. A data model consists of object types and relationship types. The metamodel resides at the highest level and is composed of metaobject types and metarelationship types. To illustrate the utility of the metamodel, the representation of the ORION data model is presented.<<ETX>>

Backend Database Systems

Backend database systems have been proposed as a solution to the problems of overloaded data processing installations. This tutorial examines backend database systems in terms of their basic structure, their potential benefits and drawbacks, and the problems facing developers of such systems. Several prototype systems are described, and research on extensions of the backend concept is discussed. The structure of the hardware and software components of backend database systems is presented in detail. The performance problems encountered in recent prototypes are pointed out and potential solutions indicated.

Data allocation in a dynamically reconfigurable environment

The authors address the data allocation problem in a dynamically reconfigurable distributed database system. Following the description of the system model and the consequences of node movement, several dynamic allocation strategies are presented. A distributed candidate-selection algorithm using the data from the accounting system in the database servers is proposed to reduce the number of files and nodes considered in reallocation. The allocation algorithms choose the optimal assignment using heuristic benefit functions and greedy search strategy. The experimental results demonstrate the accuracy and efficiency of the allocation algorithms.<<ETX>>

Extending the behavioral capabilities of the object-oriented paradigm with an active model of propagation

Object-oriented design techniques have begun to play a critical role in increasing productivity, assuring correctness, and modeling both the structure and behavior of applications. However, while most object-oriented models support intra-class behavior definition through encapsulation, inter-class relationships and behavior are supported in only a limited sense, i.e., within ISA or inheritance hierarchies. This is a serious draw-back when attempting to model advanced applications such as software-development environments (SDEs) and CAD/CAM. In order to model these critical inter-class relationships and behavior, our goal is to incorporate propagation actions into an object-oriented data model at the design-phase level. This results in increased accuracy in the modeling of information, more complete specification of system behavior, elimination of some side effects, and decreased application coding errors. To meet this goal, this paper presents the active model of propagation (AMP) which supports the specification of inter-class relationships during the design phase to permit system enforced propagation.

Towards the correctness and consistency of update semantics in semantic database schema

Discusses a paradigm and prototype system for the design-time expression, checking and automatic implementation of the semantics of database updates. Enforcement rules are viewed as the implementation of constraints and are specified, checked for consistency, and then finally mapped to object-oriented code during database design. A classification of enforcement rule types is provided as a basis for these design activities, and the general strategy for specification, analysis and implementation of these rules within a semantic modeling paradigm is discussed. SORAC (Semantics, Objects, Relationships And Constraints), a prototype database design system of the University of Rhode Island, is also described.

Representation of object-oriented data models

Abstract Database design tools have been introduced as a means of reducing the expertise required of the designer, to free him or her from the implementation details and to aid in managing the complexity of the design task. However, one of the common shortcomings of database design tools is that they are closely tied to specific data models, thus limiting the applicability of their underlaying concepts and restricting the reusability of the design software. SeaWeed, a component of the Data Model Compiler (DMC) project, proposes a solution by automatically generating database design software from data model specifications. In this paradigm, a data model specification model, or metamodel of data models, is introduced for specifying particular object-oriented data models. Using the knowledge in the specification, conceptual database design software is automatically generated for the given data model. This paper presents a metamodel for object-oriented data models and offers a technique for expressing the semantics of an object-oriented data model in terms of the metaobject types, which consist of attributes, operations, constraints and graphical representations.

A knowledge base for code reuse by similarity

A description is given of WharfRat, a knowledge base of data type implementations which employs case-based reasoning as its primary retrieval mechanism. Given a description of an abstract data type, it retrieves the most similar data type implementation in the knowledge base. The focus of the study is the process by which two case descriptions are compared. Similarity between data types is modeled by a fuzzy relation. A set of similarity matching rules has been developed and implemented. The system employs a general, graph-based data model in which object types are organized in a specialization network. Abstract data representations are built using the constructs of the general data model. This system is the first step toward developing a complete programming-by-similarity system.<<ETX>>

Reasoning by analogy in knowledge base systems

A discussion is presented of the integration of analogical reasoning into a knowledge base system. Such a system would retrieve stored examples on the basis of similarity to a current problem. A data model that permits a case-oriented representation is developed. The cases are organized in memory using an indexing scheme that permits directed search. The search strategy is based on the goal of the problem. Incorporating this type of reasoning into a database will provide a powerful search strategy which does not require users to be familiar with the organization of the database.<<ETX>>

The provision of induction in data model systems: I. Analogy

Abstract Hartzband [1] and Hartzband and Maryanski [2] have proposed that the foundation of knowledge-based systems must be an underlying compositional data model—that is, a data model that provides an object structure (the set of representational primitives supplied), an operator structure, and an inference structure (rules and relationships that control the behavior of the system). To be effective, this data model should be isomorphic with the users perception of the representation of the information. In addition, these workers have proposed that two functions that must be provided by such a system are representation of complex and/or abstract information as well as data values, and the ability to make nontrivial inferences using this information. One of the types of inference that is important in providing this perceptual isomorphism is analogy, or the evaluation of similarity among represented objects. This article describes a data model for knowledge-based systems. The inference structure of this model includes a reference structure that represents information for individual object instances. Analogy is computed as the evaluation of similarity among reference structures of object instances. Algorithms and examples are presented for several different variations.