| | |
Summary: Using patterns over ontology terms to detect
inconsistencies in large text collections
(extended abstract)
Christian Wartena1
, Anjo Anjewierden2
, and William van Dieten1
1
Telematica Instituut, Enschede, The Netherlands,
{Christian.Wartena,William.vanDieten}@telin.nl
2
Universiteit van Amsterdam, The Netherlands, anjo@science.uva.nl
Abstract. Inconsistencies in customer communication becomes a real threat for
organizations. We have investigated possibilities to detect possible inconsisten-
cies in the content of a large pension fund. An important step in the solution we
propose is finding all statements on a certain product property. In this paper we
show how this can be done using patterns over ontology terms.
1 Introduction
Many organizations are faced with the problem of a growing amount of texts that have to
be managed. One problem that arises is that inconsistencies can emerge more easily as
the amount of texts grows and constantly changes. Inconsistencies can have far reaching
|
