Sebastian Herwig

Pattern Specification and Matching in Conceptual Models - A Generic Approach Based on Set Operations

Searching for patterns in conceptual models is useful for a number of purposes, for example revealing syntactical errors, model comparison, and identification of business process improvement potentials. In this contribution, we introduce a formal approach for the specification and matching of structural patterns in conceptual models. Unlike existing approaches, we do not focus on a certain application problem or a specific modelling language. Instead, our approach is generic making it applicable for any pattern matching purpose and most conceptual modelling languages. In order to build sets representing structural model patterns, we define formal operations based on set theory, which can be applied to arbitrary models represented by sets. The basic sets represent the model elements, which in turn originate from the modelling language specification’s instances. Besides a conceptual and formal specification of our approach, we present particular application examples and a prototypical modelling tool showing its general applicability.

Formalizing Linguistic Conventions for Conceptual Models

A precondition for the appropriate analysis of conceptual models is not only their syntactic correctness but also their semantic comparability. Assuring comparability is challenging especially when models are developed by different persons. Empirical studies show that such models can vary heavily, especially in model element naming, even if they express the same issue. In contrast to most ontology-driven approaches proposing the resolution of these differences ex-post, we introduce an approach that avoids naming differences in conceptual models already during modeling. Therefore we formalize naming conventions combining domain thesauri and phrase structures based on a lin-guistic grammar. This allows for guiding modelers automatically during the modeling process using standardized labels for model elements. Our approach is generic, making it applicable for any modeling language.

Supporting Distributed Conceptual Modelling through Naming Conventions - A Tool-based Linguistic Approach

Empirical studies attest that conceptual models created in distributed modelling environments often vary heavily in the way their respective model elements are labelled. Although the same issues are being modelled, different names are chosen by the involved persons. By this, the analysis and comparison of the models, which is required for their subsequent integration, is extremely challenging and time consuming. Literature analysis reveals several approaches addressing this problem by either manually or semi-automatically integrating existing models after their construction. However, this proves to be an exhaustive and error prone task. In this article we propose a domain and modelling language independent approach that prevents the emergence of naming conflicts already during the modelling process. This is done by formalising naming conventions consisting of context specific thesauri and customised phrase structures, which are both derived from natural language grammars and supplemented by domain-specific terms. These conventions serve as basis for a fully automated guidance of the modeller during the model creation process, resulting in semantically comparable conceptual models. For this, we present a research prototype that integrates our approach into a modelling tool.

A Generic Set Theory-Based Pattern Matching Approach for the Analysis of Conceptual Models

Recognizing patterns in conceptual models is useful for a number of purposes, like revealing syntactical errors, model comparison, and identification of business process improvement potentials. In this contribution, we introduce an approach for the specification and matching of structural patterns in conceptual models. Unlike existing approaches, we do not focus on a certain application problem or a specific modeling language. Instead, our approach is generic making it applicable for any pattern matching purpose and any conceptual modeling language. In order to build sets representing structural model patterns, we define operations based on set theory, which can be applied to arbitrary sets of model elements and relationships. Besides a conceptual specification of our approach, we present a prototypical modeling tool that shows its applicability.

Enabling Widespread Configuration of Conceptual Models - An XML Approach

The manual adaptation of conceptual models in general and reference models in particular is a time consuming and error prone task, which has to be carefully conducted. The configurative reference modeling approach promises support for the model developer as well as for the model user, as certain parts of a model can be automatically removed with respect to the requirements of a certain perspective. By this automation, the risk of creating faulty models during adaptation can be highly reduced. However, up to now this approach exists in terms of its conceptual specification, leaving software support to a proprietary prototype, not providing support for widespread modeling tools, which is necessary for acceptance and applicability of the approach. To face this gap, our approach proposes an XML schema, enabling the configuration of serialized conceptual model data of virtually any modeling language and any modeling tool capable of XML export.

Forschungsinformationssysteme in Hochschulen und Forschungseinrichtungen. Positionspapier

Forschungsberichterstattung ist mittlerweile Alltag auf allen Ebenen des Wissenschaftsbetriebs:Forschende erfullen regelmasig Berichtspflichten gegenuber externen Forderorganisationen und Geldgebern. Leitungen benotigen im Wettbewerb um Ausstattung und Kopfe eine Ubersicht uber die Vielzahl an Forschungsinformationen, um gute Entscheidungen treffen zu konnen. Die offentliche Rechenschaftslegung uber Forschung, insbesondere uber die Finanzierung, hat ebenfalls an Bedeutung gewonnen. Gleichzeitig haben Hochschulen und Forschungseinrichtungen immer noch uberraschend grose Probleme bei der Bereitstellung fachbezogener Informationen. Die Ursachen sind in den Institutionen meist ahnlich – eine verteile Datenhaltung ohne Schnittstellen, Verwaltungssysteme, die Forschungskontexte nicht abbilden, sowie die begrenzte Nutzbarkeit vorhandener Systeme fur die Durchfuhrung differenzierter Auswertungen. Mehrfacherfassungen sind an der Tagesordnung. Institutionsubergreifend bestehen weitere Hindernisse fur ein effizientes Informationsmanagement: Fach- und Forderdatenbanken werden isoliert betrieben, es fehlt an Schnittstellen und Austauschformaten und bei der Entwicklung von Systemen werden Standardisierungsmoglichkeiten zu wenig genutzt. Die Entwicklung finanzierbarer und funktionaler Forschungs-informationssysteme und vor allem der Austausch vorhandener Informationen gehort ebenso zur IT-Entwicklung fur die Wissenschaft, wie Campusmanagement oder geeignete Personal-und Finanzsysteme. Es ist schwer vorstellbar, dass Prozesse, die auf manuelle Eingaben und jahrliche Datenabfragen angewiesen sind, in den zahlreichen wissenschaftlichen Einrichtungen dieser Welt auf Dauer erfolgreich fortgefuhrt werden konnen. Ebenso wenig ist wahrscheinlich, dass die Berichtsanforderungen abnehmen. Das Positionspapier zeigt auf strategischer Ebene auf, welche konkreten Schritte zur Entwicklung eines nachhaltigen Informationsmanagements in der Forschungsberichterstattung gegangen werden mussen. Eine Verstandigung auf gemeinsame Standards ist die Grundvoraussetzung dafur, dass der nicht unbetrachtliche Aufwand fur den Betrieb der Systeme gesenkt werden kann und mobilen Forschenden ermoglicht wird, ihr Portfolio in verschiedene Anwendungen zu ubertragen und auch in andere Forschungseinrichtungen mitzunehmen. Die Arbeitsgruppe hat daruber hinaus konkrete handlungspraktische Hinweise fur Systemauswahl, Einfuhrung und Betrieb, Projektmanagement und Gestaltung der Systeme erarbeitet. Diese richten sich an Institutionen, die selbst ein Forschungsinformationssystem einfuhren oder entwickeln wollen.