Please use this identifier to cite or link to this item:
Title: Ontology-based author profiling of documents
Authors: Jarrar, Mustafa
De Bo, Jan
Majer, Ben
Keywords: System design;Expert systems (Computer science);Artificial intelligence
Issue Date: 2014
Abstract: In this paper we present the advantages of using an ontology service for the modelling of user profiles in the EC FP5 IST project NAMIC (IST-1999-12392). By means of an ontology server people set up user profiles, which are in fact views, i.e. specifications of queries on the ontology. These views are constructed using a JAVA API, which forms the commitment layer of the ontology, built on top of an ontology base. In NAMIC an ontology server is used to establish a link between the lexical object representations, generated by the natural language processors (NLP) on the one hand and the user’s interest, specified through the selection of relevant concepts and facts of the ontology on the other. This allows to specify a user profile independently of language, categorization and NLP specific "world models". Users then set up a profile consisting of events, agents participating in these events and other content information in which they are interested in. For instance, a journalist writing articles about financial issues may be interested in related documents containing a “raise event” of company shares. If he has specified those conditions in his profile he will be able to retrieve resources which contain events that are semantically related to that kind of event pattern. User profiles in NAMIC do not have to be static. The results of processing by the NLPs of a document the user is currently working on, may be used to construct a dynamic profile, which may contain events specific for that document. This way a user’s profile can be dynamically adapted to his current interests. We also developed a tool which illustrates the creation of user profiles using ontological concepts and facts
Appears in Collections:Fulltext Publications

Files in This Item:
File Description SizeFormat
Ontology_based_author_profiling_of_documents_ LREC_Published.pdf433.34 kBAdobe PDFView/Open
Show full item record

Page view(s)

Last Week
Last month
checked on Jun 27, 2024


checked on Jun 27, 2024

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.