Arne Sølvberg

Understanding quality in conceptual modeling

With the increasing focus on early development as a major factor in determining overall quality, many researchers are trying to define what makes a good conceptual model. However, existing frameworks often do little more than list desirable properties. The authors examine attempts to define quality as it relates to conceptual models and propose their own framework, which includes a systematic approach to identifying quality-improvement goals and the means to achieve them. The framework has two unique features: it distinguishes between goals and means by separating what you are trying to achieve in conceptual modeling from how to achieve it (it has been made so that the goals are more realistic by introducing the notion of feasibility); and it is closely linked to linguistic concepts because modeling is essentially making statements in some language.<<ETX>>

Evaluating the quality of information models: empirical testing of a conceptual model quality framework

This paper conducts an empirical analysis of a semiotics-based quality framework for quality assuring information models. 192 participants were trained in the concepts of the quality framework, and used it to evaluate models represented in an extended Entity Relationship (ER) language. A randomised, double-blind design was used, in which each participant independently reviewed multiple models and each model was evaluated by multiple reviewers. A combination of quantitative and qualitative analysis techniques were used to evaluate the results, including reliability analysis, validity analysis, interaction analysis, influence analysis, defect pattern analysis and task accuracy analysis.. An analysis was also conducted of the frameworks likelihood of adoption in practice. The study provides strong support for the validity of the framework and suggests that it is likely to be adopted in practice, but raises questions about its reliability and the ability of participants to use it to accurately identify defects. The research findings provide clear directions for improvement of the framework. The research methodology used provides a general approach to empirical validation of quality frameworks.

Information Systems Engineering

ion is a useful mechanism in conceptual modeling. Abstractions can be achieved in semantic networks by using spaces and compound spaces. The central idea is to allow groups of nodes and arcs to be bundled together 488 14. Fonnal Modeling Approaches

Evaluating the Quality of Process Models: Empirical Testing of a Quality Framework

This paper conducts an empirical analysis of a conceptual model quality framework for evaluating the quality of process models. 194 participants were trained in the concepts of the quality framework, and then used it to evaluate models represented in a workflow modelling language. A randomised, double-blind design was used, and the results evaluated using a combination of quantitative and qualitative techniques. An analysis was also conducted of the frameworks likelihood of adoption in practice, which is an issue rarely addressed in IS design research. The study provides strong support for the validity of the framework and suggests that it is likely to be adopted in practice, but raises questions about its reliability. The research findings provide clear direction for further research to improve the framework.

Data and What They Refer to

In data modeling there is an implicit assumption of a one-to-one correspondence between a data model and the world which the data convey information about. Each data item (value) in the data base corresponds to a property of the world. Our view of the data base reflects our view of the world. For simple situations this is enough, e.g., one data record for each person. For complex situations the simple one-to-one correspondence is no longer enough. When the number of worldly phenomena and the number of data names grow, it becomes increasingly difficult to keep track of how the various concepts relate to each other. Many different views of the world may co-exist, each view serving different purpose and/or different people. No view is more correct than another because each view serves a worthy purpose. Conceptual data models have been proposed as tools for relating the various world views. For many years research has been conducted in the data base field, in artificial intelligence and information systems to find representions of knowledge that may be easily accepted among users as well as among designers of software and data bases. In spite of the many research efforts we find that various dialects of the ER-model still dominate in the practical world. For a conceptual model to be successful it should relate well to common sense views of the world, and also relate well to commonly known mathematical formalisms. These are necessary preconditions for being widely accepted. We propose a modeling framework which meets these two conditions. We relate to a common sense view of the world which is based on the old distinctions among ideas, concepts, matter, and images. The mathematical form is elementary discrete mathematics, which in its most simple form is common sense knowledge, even known to children in the elementary school. A visual language for data modeling is supported by a tool which at present consists of an editor and an administrative system for supporting concurrent information systems engineering.

Semantic annotation framework to manage semantic heterogeneity of process models

Effective discovery and sharing of process models within and/or across enterprises are important in process model management. A semantic annotation approach has been applied for specifying process semantic heterogeneity in the semantic process model discovery in our previous work. In this paper, the approach is further developed into a complete and systematic semantic annotation framework. Four perspectives are tackled in our framework: basic description of process models (profile annotation), process modeling languages (meta-model annotation), process models (model annotation) and the purpose of the process models (goal annotation). Ontologies, including modeling ontology, domain specific ontology and goal ontology, are used for annotation of process models to achieve semantic interoperability. A set of mapping strategies are defined to guide users to annotate process models.

Goal annotation of process models for semantic enrichment of process knowledge

A semantic annotation framework has been proposed to tackle the semantic heterogeneity problem of distributed process models in our earlier work. The goal annotation as part of the framework is further developed, in which goal ontology is annotated to process models to indicate the objectives or the capability of models. In the paper, we introduce a way to represent goal ontology, build relationships between goals and process models, and develop a goal annotation approach to process models. As an illustration, a case study is deployed with the proposed annotation approach. The results of the goal annotation enrich the semantics of process knowledge from stakeholders perspective in a cooperative goal-oriented manner. The ontology and the annotation results also facilitate the ontology-based queries for the semantic discovery and the reuse of heterogenous process models.

Project Courses at the NTH: 20 Years of Experience

Project courses are a cornerstone in the information systems and software engineering education offered at the NTH (Norges Tekniske Hogskole = Norwegian Institute of Technology). In this paper we present our two main project courses, one taught in the 2nd year and one in the 4th year.

A framework for performance engineering during information system development

Software performance engineering aims at predicting and improving the performance of applications during development and in production. This paper presents a framework for performance engineering of information systems with emphasis on parameter estimation support.

A composite modelling approach to software performance measurement

Traditionally performance modellers have tended to ignore the difficulty of obtaining parameter vaules which represent the resource demands of multi-layered software. In practice the use of performance engineering in large-scale systems development is limited by the cost of acquiring appropriate performance information about the various software components. However, if this information cart be reused when components are combined in different ways, then the cost of measurement can be more easily justified. Such reuse can be achieved by means of a composite work model.