Semantic decision table

A semantic decision table uses modern ontology engineering technologies to enhance traditional a decision table. The term "semantic decision table" was coined by Yan Tang and Prof. Robert Meersman from VUB STARLab (Free University of Brussels) in 2006. ^[1] A semantic decision table is a set of decision tables properly annotated with an ontology. It provides a means to capture and examine decision makers’ concepts, as well as a tool for refining their decision knowledge and facilitating knowledge sharing in a scalable manner.

Background

A decision table is defined as a "tabular method of showing the relationship between a series of conditions and the resultant actions to be executed". ^[2] Following the de facto international standard (CSA, 1970), a decision table contains three building blocks: the conditions, the actions (or decisions), and the rules.

A decision condition is constructed with a condition stub and a condition entry. A condition stub is declared as a statement of a condition. A condition entry provides a value assigned to the condition stub. Similarly, an action (or decision) composes two elements: an action stub and an action entry. One states an action with an action stub. An action entry specifies whether (or in what order) the action is to be performed.

A decision table separates the data (that is the condition entries and decision/action entries) from the decision templates (that are the condition stubs, decision/action stubs, and the relations between them). Or rather, a decision table can be a tabular result of its meta-rules.

Traditional decision tables have many advantages compared to other decision support manners, such as if-then-else programming statements, decision trees and Bayesian networks. A traditional decision table is compact and easily understandable. However, it still has several limitations. For instance, a decision table often faces the problems of conceptual ambiguity and conceptual duplication^{[ citation needed ]}; and it is time consuming to create and maintain large decision tables^{[ citation needed ]}. Semantic decision tables are an attempt to solve these problems.

Definition

A semantic decision table is modeled based on the framework of Developing Ontology-Grounded Methods and Applications (DOGMA ^[3] ). The separation of an ontology into extremely simple linguistic structures (also known as lexons) and a layer of lexon constraints used by applications (also known as ontological commitments), aiming to achieve a degree of scalability.

According to the DOGMA framework, a semantic decision table consists of a layer of the decision binary fact types called semantic decision table lexons and a semantic decision table commitment layer that consists of the constraints and axioms of these fact types.

A lexon l is a quintuple ${\displaystyle <y,t_{1},r_{1},r_{2},t_{1}>}$ where ${\displaystyle t_{1}}$ and ${\displaystyle t_{2}}$ represent two concepts in a natural language (e.g., English); ${\displaystyle r_{1}}$ and ${\displaystyle r_{2}}$ (in, ${\displaystyle r_{1}}$ corresponds to "role and ${\displaystyle r_{2}}$ – refer to the relationships that the concepts share with respect to one another; ${\displaystyle \gamma }$ is a context identifier refers to a context, which serves to disambiguate the terms ${\displaystyle t_{1},t_{2}}$ into the intended concepts, and in which they become meaningful.

For example, a lexon <γ, driver's license, is issued to, has, driver> explains a fact that “a driver’s license is issued to a driver”, and “a driver has a driver’s license”.

The ontological commitment layer formally defines selected rules and constraints by which an application (or "agent") may make use of lexons. A commitment can contain various constraints, rules and axiomatized binary facts based on needs. It can be modeled in different modeling tools, such as object-role modeling, conceptual graph, and Unified Modeling Language.

Semantic decision table model

A semantic decision table contains richer decision rules than a decision table. During the annotation process, the decision makers need to specify all the implicit rules, including the hidden decision rules and the meta-rules of a set of decision tables. The semantics of these rules is derived from an agreement between the decision makers observing the real-world decision problems. The process of capturing semantics within a community is a process of knowledge acquisition.

Notes

1. Yan Tang & Robert Meersman (2007). C. Man-chung; J.N.K. Liu; R. Cheung & J.Zhou (eds.). Towards building semantic decision table with domain ontologies. Proceedings of the International Conference of Information Technology and Management (ICITM2007). ISM Press. pp. 14–21. ISBN   978-988-97311-5-1.
2. Canadian Standards Association (1970). Z243.1–1970 for Decision Tables.
3. Robert Meersman (2001). d'Atri, A.; Missikoff, M. (eds.). Ontologies and Databases:More than a Fleeting Resemblance. Proc. of OES/SEO 2001 Rome Workshop. Luiss Publication.

References

• Canadian Standards Association (1970). Z243.1–1970 for Decision Tables.
• Yan Tang & Robert Meersman (2007). C. Man-chung; J.N.K. Liu; R. Cheung & J. Zhou (eds.). Towards building semantic decision table with domain ontologies. Proceedings of the International Conference of Information Technology and Management (ICITM2007). ISM Press. pp. 14–21. ISBN   978-988-97311-5-1.
• Yan Tang & Robert Meersman (2008). Man-Chung Chan; Ronnie Cheung & James N K Liu (eds.). Towards Building Semantic Decision Tables with Domain Ontologies. Challenges in Information Technology Management. World Scientific. ISBN   978-981-281-906-2.
• Yan Tang, Robert Meersman and Jan Vanthienen. S. Bhwmich; Josef Kung; Roland Wagner (eds.). Semantic Decision Tables: Self-Organizing and Reorganizable Decision Tables. Proceedings of DEXA'08 (19th International Conference on Database and Expert Systems Applications). Turin, Italy: Springer. LNCS 5181.
• Yan Tang & Robert Meersman (2009). "Use Semantic Decision Tables to Improve Meaning Evolution Support Systems". In Frode Eika Sandnes; Yan Zhang; Chunming Rong; Laurence T. Yang; Jianhua Ma; et al. (eds.). International Conference on Ubiquitous Intelligence and Computing. doi:10.1007/978-3-540-69293-5_15. ISBN   978-3-540-69293-5.
• Yan Tang & Robert Meersman (2009). SDRule Markup Language: Towards Modelling and Interchanging Ontological Commitments for Semantic Decision Making. Handbook of Research on Emerging Rule-Based Languages and Technologies: Open Solutions and Approaches. IGI Publishing, USA. ISBN   978-1-60566-402-6.