209 
SPATIAL DATA UNCERTAINTY MANAGEMENT 
D. Klimesovâ, E. Ocelikovâ* 
Czech University of Life Sciences, Prague, Faculty of Economics and Management, Kamÿckâ 129, 165 21 
Prague 6, Czech Republic - klimesova@pef.czu.cz 
Czech Academy of Sciences, Institute of Information Theory and Automation, Pod vodârenskou vëzi 4,182 00 
Prague 8, CR - klimes@utia.cas.cz 
technical University Kosice, Faculty of Electrical Engineering and Informatics, Department of Cybernetics 
and Artificial Intelligence, Letnâ 9, 042 00 Kosice, Slovak Republic - ocelike@ccsun.tuke.sk 
Commission IV, WG IV/2 
KEY WORDS: Knowledge, Classification, Knowledge Management, Contextual Modelling, Temporal Modelling, Decision 
Support 
ABSTRACT: 
The paper deals with the relations between knowledge management, uncertainty and the context evaluation on the background of the 
new possibilities of information technologies that can help us to carry out the knowledge management strategies. The paper discusses 
the problem of wide context including to compensate and decrease the uncertainty of data on one hand and to increase accuracy of 
classification or segmentation and efficiency of further analytical processes to increase the information value of decision support. The 
technique of fuzzy measurements and a fuzzy approach in general is addressed also as a way to catch the uncertainties of analytical 
process (classification, segmentation, ...) and to transmit them to the other processing stages. 
1. INTRODUCTION 
The ways of managing and distributing data and particularly 
data sources has rapidly changed. Data are collected and 
processed and during the last couple of years the data flows in 
and between organizations have extremely increased. In the 
connection with these facts also the data management tools and 
techniques are continually changed [10]. 
Knowledge management (KM) is an essential process 
improving competitive advantage. KM tries to compensate the 
loss of stable procedural knowledge, the loss of customer- 
related or project-related experiences and know-how or the loss 
of the middle management information analysis and routing 
services. 
The area in Computer Science that is most influenced by the 
concept of knowledge is Artificial Intelligence and Knowledge 
Based Systems together with Geographic Information Systems 
(GIS) [6]. Since the knowledge is specified independently from 
the application domain, reuse of the knowledge is enabled for 
different domains and applications [5]. 
The modelling process is dependent on the subjective 
interpretation of the knowledge engineer. Therefore this process 
is faulty and an evaluation of the model with respect to reality is 
indispensable for the creation of an adequate model. Since this 
control knowledge is specified independently from the 
application domain, reuse of this strategically knowledge is 
enabled for different domains and applications. Besides 
knowledge modelling also knowledge representation is an 
important field of research in computer science and AI. 
Besides knowledge modelling and knowledge representation is 
also an important field of the building process of knowledge 
database. The modelling process is a cyclic process that may 
lead to the refinement, modification, or completion of the 
already constructed model or guide further acquisition of 
knowledge - contextual understanding. 
2 UNCERTAINTY 
2.1 Multi-source Data 
The Spatial variability, the dimensionality of data and the 
complexity of objects structure hierarchy are rapidly growing 
and consequently with these aspects increase the uncertainty 
entering into the processing. A great number of existing 
databases offer a variety of data sets covering different thematic 
aspects like topographic information, cadastral data, statistical 
data, digital maps, aerial and satellite images including temporal 
data. 
GIS works with the combination of data sets which may have a 
very different uncertainty structure. There is the question of 
correlation and the necessity to consider the covariances 
together with variances. 
The problem of scale exists - the loss of knowledge about the 
variability within the map unit, the quality of estimations and 
the unknown model of overall uncertainty. 
GIS is proposed as multisource technology as opened system 
with wide spectra of interconnections. But on the other hand It 
is quite another situation than in case of the closed system, 
where the user has full control over all steps of processing from 
data input to presentation of results. In frame of open 
interoperable system with access to web sources with a great 
number of existing databases the user control gets completely 
lost. User needs an appropriate uncertainty model for this 
purpose, integrated in GIS [4]. 
Uncertainty arises from imperfect understanding of the events 
and processes. From the philosophical point of view the