Arno Siebes

Visualizing association rules with interactive mosaic plots

Association rules are amongst the most important patterns that can be discovered using data mining. Their automatic discovery is supported by most, if not all, data mining software tools. Moreover, many techniques have been devised to lter the most interesting (in many senses) rules from the complete set of discovered rules, such that the users are not swamped by results. However, association rules are actually hard to understand; even more so if one only looks at the most interesting rules. For example, rather strong correlations between attributes are not always obvious from the discovered rules. Similarly, a deeper explanation of related association rules may be missing from the rule set. In this paper we show how Mosaic plots and, especially, their variant called Double Decker plots, can be used to visualize association rules. These plots visualize the contingency table that yields the association rule as well as the other potential rules in that table, whether they meet the thresholds or not. This gives a deeper understanding on the nature of the correlation between the left-hand side of the rule and the right-hand side. Moreover, we show how an interactive use of these plots helps the user to understand the relationship between related association rules.

Schema Integration in Object-Oriented Databases

This paper presents a formal approach to support schema integration in object-oriented databases. The basis of the approach is a subclass order, which is defined in terms of a weak subtype relation on underlying types of classes and a subfunction relation on functional forms of methods. The subclass order induces an equivalence relation and a join operator, which are used to identify and factorise class hierarchies, leading to a natural framework for integration of class hierarchies. The novelty of this paper is that both attributes and methods are used to compare classes, and that behaviour of methods is used to compare attributes, resulting in a more semantic approach towards schema integration in object-oriented databases.

Multi-relational Decision Tree Induction

Discovering decision trees is an important set of techniques in KDD, both because of their simple interpretation and the efficiency of their discovery. One disadvantage is that they do not take the structure of the data into account. By going from the standard single-relation approach to the multi-relational approach as in ILP this disadvantage is removed. However, the straightforward generalisation loses the efficiency. In this paper we present a framework that allows for efficient discovery of multi-relational decision trees through exploitation of domain knowledge encoded in the data model of the database.

An approach to schema integration based on transformations and behaviour

This article presents an approach to schema integration that combines structural aspects and behavioural aspects. The novelty of the approach is that it uses behavioural information to guide both schema restructuring and schema merging. Schema restructuring is based on schema transformations and schema merging is based on join operators.

Enriching Active Databases with Agent Technology

Intelligent agents are software components with a largely autonomous behaviour, that are fitted out with a considerable degree of artificial intelligence. They are a promising paradigm to serve as a foundation for future computing environments in general, and information systems in particular. At the same time database research has seen the rise of active databases, database systems that add autonomous behaviour to a database. In this paper, we investigate the addition of notions from intelligent agents to an active database. We explain why active databases already implement weak agency, and look into the benefits stronger agency can bring to an active database. It turns out that these are mainly found in the increased flexibility facilitated by the reasoning abilities strong agency implies. For example, an agent can have multiple strategies to maintain a constraint instead of a one fixed strategy defined by triggers.

Multi-Relational Data Mining, Using UML for ILP

Although there is a growing need for multi-relational data mining solutions in KDD, the use of obvious candidates from the field of Inductive Logic Programming (ILP) has been limited. In our view this is mainly due to the variation in ILP engines, especially with respect to input specification, as well as the limited attention for relational database issues. In this paper we describe an approach which uses UML as the common specification language for a large range of ILP engines. Having such a common language will enable a wide range of users, including non-experts, to model problems and apply different engines without any extra effort. The process involves transformation of UML into a language called CDBL, that is then translated to a variety of input formats for different engines.

Data Mining and the KESO Project

Data Mining and Knowledge Discovery is a young but vigorously growing research area. Its aim is to discover structure or knowledge in databases. It comprises a wide variety of algorithms and techniques for towards this goal.

DEGAS: Capturing Dynamics in Objects

In this paper we introduce Degas (Dynamic Entities Get Autonomous Status), an active temporal data model based on autonomous objects. The active dimension of Degas means that we define the behaviour of objects in terms of production rules. The temporal dimension means that the history of an object is included in the Degas data model. Novel features of Degas axe the encapsulation of the complete behaviour of an object, both potential and actual. Thus, Degas combines dynamic and structural specifications in one model. In addition, Degas allows easy evolution of object capabilities through a clear distinction between inherent types and capabilities that can be acquired and lost. This addon mechanism makes Degas very suitable as a formalism for role modelling. Finally, the rule model in Degas is both simple, through the use of finite automata, and general, because it allows different strategies for dealing with constraints and reacting to events in other objects.

Enforcing Confluence of Rule Execution

Rules provide the functionality for constraint enforcement and view maintenance. A provably correct implementation of both issues based on rules, requires confluent and terminating behaviour of the rule set. In [15], we introduced a design theory for the static detection of these properties. The detection of confluence is based on commutativity of rule execution, called independence. In this article, we discuss the enforcement of confluence for terminating, dependent rule sets.

DEGAS: a database of autonomous objects

In this paper we introduce DEGAS (Dynamic Entities Get Autonomous Status), an active temporal data model based on autonomous objects. The natural combination of active and temporal databases is discussed. The active dimension of DEGAS means that we define the behaviour of objects in terms of production rules. The temporal dimension means that the history of an object is included in the DEGAS data model. Further novel features of DEGAS are the encapsulation of the complete behaviour of an object, both potential and actual. Thus, DEGAS combines dynamic and structural specifications in one model. In addition, DEGAS allows easy evolution of object capabilities through a clear distinction between inherent types and capabilities that can be acquired and lost. This addon mechanism makes DEGAS very suitable as a formalism for role modelling. Finally, the rule model in DEGAS is both simple, through the use of finite automata, and general, because it allows different strategies for dealing with constraints and reacting to events in other objects.