On the conceptualisation of the query refinement task
| DOI | https://doi.org/10.1108/01435120510596125 |
| Pages | 281-294 |
| Date | 01 May 2005 |
| Published date | 01 May 2005 |
| Author | Nenad Stojanovic |
| Subject Matter | Library & information science |
On the conceptualisation of the
query refinement task
Nenad Stojanovic
Institute AIFB, University of Karlsruhe, Karlsruhe, Germany
Abstract
Purpose – To resolve the drawback in the existing methods for query refinement by using semantic
technologies.
Design/methodology/approach – The article introduces a conceptual space (ontology) in which a
query refinement problem will be treated. In that space a query is modeled as a set of logic formulas,
query refinement process is modeled as an inferencing process. The approach is focused on the
positive conjunctive queries.
Findings – This (re)opens a question about a need for a more semantic-based indexing of documents
in the traditional information retrieval in order to improve the overall retrieval performances. This
experimental study has shown that the semantics arisen (emerged) from row documents has enough
quality to enable building query refinement methods that outperform the most advanced
syntactic-driven methods. Consequently, one can imagine a more em ergent-semantic based
indexing process that will retain the autonomy of web data, but will significantly increase the
precision of the retrieval process performed by a traditional search engine.
Originality/value – A novel approach for the query refinement is presented. It deals with the
conceptual level of a user’s query. Another benefits is that this conceptual structure emerges from the
textual data automatically. The case study shows how a traditional full-text search engine can benefit
from applying this approach.
Keywords Information searches, Document management, Indexing, Information retrieval
Paper type Research paper
Introduction
A very important characteristic of an information retrieval process is its exploratory
nature, whereby:
.A user is usually unfamiliar with the content of the information repositories. In
order to avoid making an over-specified “long” query that retrieves zero results
(i.e. a failing query), the user starts searching with a short query and tries to
exploit the repository in several subsequent refinement steps.
.A user often has ill-defined information need. He starts searching by assuming
what can be the right information, but often, by exploring the resulting list, he
redefines what he is actually searching for.
In the nutshell of this exploratory process is the process of expanding or redefining the
initial query in order to obtain more relevant results – the so-called query refinement
process. However, existing methods for query refinement seem to be inadequate for
real-world usage (Campbell, 2000), since they usually return a long list of refinements,
which is hard to process manually. Indeed, the recent experimental studies in the
The Emerald Research Register for this journal is available at The current issue and full text archive of this journal is available at
www.emeraldinsight.com/researchregister www.emeraldinsight.com/0143-5124.htm
Research for this paper was partially financed by BMBF in the project “SemIPort” (08C5939) and
EU in the project “KnowledgeWeb” (507482).
Query
refinement task
281
Library Management
Vol. 26 No. 4/5, 2005
pp. 281-293
qEmerald Group Publishing Limited
0143-5124
DOI 10.1108/01435120510596125
To continue reading
Request your trial