Knowledge Discovery for Interactive Dialogue Management with

Geoinformation Service

Stanislav Belyakov, Alexandr Bozhenyuk, Marina Belyakova and Sergey Zubkov

Institute of Computer Technologies and Information Security, South Federal University, Chekhov St. 2, Taganrog, Russia

Keywords: Knowledge Discovery, Visualization of Spatial Data, Geographic Information Systems, Intelligent Systems,

Adaptation.

Abstract: The problem of adaptation of the geoinformation service to the increasing amount of knowledge and

modification of the spatial database structure is analyzed in this article. The necessity to take into account

the factors of changes in information basis is considered by a visualization of searching procedure and by

analyzing spatial data of the geoinformation service. The paper proposes a method for solving the problem,

based on the principle of the evolution of technical systems. In this problem the evolutionary principle

considers the continuous rules generation procedure by the geoinformation service containing knowledge of

useful cartographic objects for visual analysis. The rules are considered as hypotheses that require collective

confirmation from the clients of the service. Confirmation of any rule is a selection of useful knowledge for

further implementation. Thus, the proposed mechanism provides a continuous adaptation to a changing

information environment through the development and rule selection. The mechanism of generation is

analyzed and the structure of rules is determined. The mechanism of collective confirmation of rules is

considered as well.

1 INTRODUCTION

Network geoinformation services are systems for the

collective use of spatial data. The geoservice

information database includes charts, maps, plans,

earth surface photographs and other materials that

reflect real-world phenomena and objects (Shashi

and Hui, 2008). Depending on the field of

application, geoservices have different sizes - from

corporate to geoservices in Internet (Brimicombe

and Chao, 2009). Geoservice of any size (scale) can

be attributed to BigData systems . This feature is

determined by the (appointment of any geographic

information system, which consists in the continuous

accumulation of data on the changing world around.

The volume of spatial data is constantly growing,

existing structures and their representations are

changing (Yang at al., 2015), and new structures are

emerging. Spatial data never become obsolete, only

the relevance is changed with regard to the solution

of specific applied problems. The use of spatial data

is impossible without a set of special methods aimed

at obtaining information useful for the solution of

the problem. It is of great interest to develop some

methods that can be used for visual representation

and visual analysis of spatial data. Cartographic

visualization is the most effective tool for solving

complex informal problems. Software tools of

special types of analysis (statistical, topological,

morphological, network and many others) play an

auxiliary role. They complement the virtual

cartographic image that the map analyst develops

(Belyakov et al., 2014).

The selection of useful information largely

determines the final result of spatial data usage. In

accordance with the principles of cartographic

research (Shashi and Hui, 2008) the user should

build a workspace of the geoservice information

base. The workspace includes cartographic objects

that are grouped into thematic layers and linked by

links to external databases. The problem complexity

of constructing a workspace is to select useful data

from the BigData source. This source is the

geospatial database. Unsatisfactory implementation

of this problem entails losses caused by the adoption

of inadequate decisions.

Let's consider a simple example. When planning

the placement of a vending machine on the territory

of a residential complex, the analyst is looking for

cartographic data that to help evaluate the

Belyakov, S., Bozhenyuk, A., Belyakova, M. and Zubkov, S.

Knowledge Discovery for Interactive Dialogue Management with Geoinformation Service.

DOI: 10.5220/0006836504350442

In Proceedings of 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2018), pages 435-442

ISBN: 978-989-758-323-0

435

effectiveness of the decision. The intensity of the

flow of people passing by the machines, the

approximate share of potential buyers, the relative

position of objects or phenomena, the size of the

access zone for maintenance of the machine, the

logistics capabilities of the placement point from the

perspective of further business development have a

complex effect on the efficiency of choice. In Fig. 1

shows the section of the working area of the

analysis, which was formed on the basis of

experience in solving the type of problems under

consideration. The selected position of the vending

machine is marked with the special sign. In Fig. 2

shows the important elements for analysis, which

one of the analysts used to construct the solution.

These elements are data on the density of cars in

parking lots along highways. Since a number of

standing cars greatly reduces the visibility of the

vending machine by nearby people, this decision

may be erroneous. Analyzing this example, you can

see a potential opportunity to improve the quality of

geoservice: one of its users found useful data, which,

perhaps, can be successfully used by other analysts

who solve similar tasks.

Figure 1: Example of the working area.

Figure 2: The important elements for analysis.

In this paper, an approach based on knowledge

discovery is analyzed to improve the efficiency of

visual analysis of maps and charts in an interactive

mode of working with geoservice.

2 APPROACHES OVERVIEW

There are several approaches to solving the problem

of selecting spatial data useful for analysis. The

main idea of each approach is a special concept

definition for the search and use of knowledge,

which improves the quality of visual analysis of

geodata.

The methodology of traditional cartographic

research (Shashi and Hui, 2008) assumes the

creation of a thematic map for solving a certain class

of applied problems. The professional knowledge

and experience of the cartographer, as well as

fundamental knowledge of cartography, play a

decisive role in this case. The search and

generalization of maps of different subjects have a

pronounced creative character. The result is unique

cartographic works. The nature of this activity is

very different from the procedures for searching and

analyzing maps for solving specific applications.

Therefore, the methodology adopted in the field of

the search for and use of knowledge cannot be

directly used.

Works on geo-visual analytics (Andrienko et al.,

2013) examine the mechanisms for the formation of

new cartographic representations that are adequate

to the goals of analysis. The significance of a new

view is determined by new knowledge and patterns

that the analyst can identify. Software tools play an

important role in scaling, aggregating and

summarizing cartographic images. It can be

concluded that the selection of abstractions and the

development of tools for visual analysis of the

generated image form the basis of the approach

under consideration. So it is possible to obtain

practically important conclusions and dependences (

Andrienko and Andrienko, 2016). At the same time,

manipulating the cartographic image for presentation

system is a more general process. Perception of

images by the user plays a decisive role in any way

of displaying. The problem of reducing the utility of

cartographic visualization with its complication in

the work of this direction is not considered.

Neo-cartography (Faby and Koch, 2010) aims to

display the surrounding world in real time. This

approach is focused on presenting the real world

with multimedia data. The user interface is

intentionally simplified, which is important for

geoinformation Internet services. A particular

feature of this approach is the weakening of the

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

436

connection with such important mechanisms for the

synthesis of cartographic images as cartographic

classification and generalization. Accordingly,

interactive interaction has a specific semantic

orientation, which makes it difficult to use the

regularities of dialogue in traditionally constructed

geoservices.

Research in the field of intellectual visualization

(Pettit et al., 2008; Malczewski, 2004; Keim, 2002)

aims to create procedures for the selection of useful

information. As the analysis has shown, this problem

is solved by the creation of complete ready-to-use

information resources. At the same time, the

problem of identifying knowledge remains little

explored for constructing universal mechanisms for

selection of thematic data in local areas of spatial

databases.

The idea of collective usage of spatial data is

studied in the framework of the direction of the

creation of the Global Spatial Data Infrastructure (Li

et al., 2012; Wang 2010). Here the main attention is

paid to the technology of sharing spatial data and

servicing the BigData data storage system. At the

same time, the problems of discovering the patterns

of using information in a dynamic database

structure, the collective experience of rational

selection of data remain outside the scope of

research.

The research direction related to ensuring the

quality of geodata, covering all stages of obtaining

and using spatial data, is, from our point of view, the

most promising (Hart and Dolbear, 2013; Ali and

Schmid, 2014; Goodchild and Li, 2012; Fairbairn,

2015). Identifying and using knowledge to manage

the quality of workspaces and solving application

classes (Popovich et al., 2011) is an important but

insufficiently studied problem.

The following works (Lughofer, 2011; Angelov,

Filev and Kasabov, 2010; Angelov, 2012) are

devoted to developing (evolving) systems. The

mechanism of evolution suggests the accumulation

of knowledge in real-time and automatic learning.

This direction of building systems allows to realize

autonomous adaptation to changing external

conditions. The results of studies deal with the

processing of numerical data streams and fuzzy rules

of classification. It seems appropriate to extend this

mechanism to the process of dialogue between

clients and geoservice.

3 THE PRINCIPLE OF

OPERATION GEOSERVICE

Geoservice is an intelligent information system that

serves analyst users through software clients. Clients

realize the functions of visualization of cartographic

data obtained at the request of geoservice. The

database of cartographic data and knowledge is

stored on the server.

Consider in general the task of optimizing the

dialogue that geoservice decides. Let

},...,{

21 n





be the set of cartographic objects

of the spatial database, and

w

is the workspace

of the map created by the user to find the solution to

the applied task. The complexity of the workspace

(the number of cartographic objects in it) is usually

much less than the complexity of the service

database:

.|||| w

This ratio is achieved as a result of the user

performing a sufficiently large amount of work on

the selection of useful information. The reason for

this is a significant redundancy in the spatial data

stream generated by spatial queries (Shashi and Hui,

2008).

We denote by

)(wI

the utility function of the

workspace and by

qqq ,...,

- the sequence of the

software client's requests to the geoservice. Then

geoservice must solve the problem:





















EBw

max,)(



(1)

here



is the set of cartographic objects

received by the software client on request of

;



is a set of cartographic objects selected

according to the

rule from the geoservice

knowledge base of the

),( BR

. The

(workspace

skeleton) set is a set of objects formed by the server

at the client's request, and the

set is the set of

objects that are the semantic addition of the skeleton.

),(  BRr

rules define those cartographic objects

that are selected from the spatial database and are

entered into a workspace built from a number of

cartographic objects of the skeleton

. Knowledge

Knowledge Discovery for Interactive Dialogue Management with Geoinformation Service

437

can be applied to an arbitrary set of objects. The

result will be an image of reality in the context of the

knowledge used. It should be noted that the utility

)(wI

of the workspace is changed by varying the

environment. For example, a car route constructed

on request is presented in different ways in the

context of transport of people or large-sized cargoes.

As shown in (Belyakov et al., 2014), the solution

of the utility maximization problem is to find the

classes, objects and relations that make up the set of

the environment

, and specify a preference

relation that allows us to establish a non-strict order

on the set of its elements.

It should be noted that

),( BR

contains

fundamental knowledge for solving an applied

problem. This knowledge is distinguished by a high

level of generalization and, for obvious reasons,

does not take into account local factors important for

mapping in specific contexts. For example, when

implementing a logistics project, it is natural to

display on the map a transport highway with the

points of loading and unloading the vehicle.

However, in some parts of the locality at certain

times of the year, unfavorable weather conditions

may occur. This will require knowledge of

additional parking spaces, transshipment or

repackaging of cargo.

Lack of special knowledge is an objective

property of the geoservice knowledge base. This

property will be preserved, despite any attempts to

achieve completeness. The reason is the variability

of the real world. In order to compensate for the

incompleteness, it is proposed to use the possibility

of temporarily disabling the intellectual support of

the dialog in the normal operation of the service.

This will allow the analyst to block "unreasonable"

from his point of view the behavior of the system, as

well as to indicate periods of use by the analyst of

their own knowledge that improve the quality of the

work area.

Geoservice, which solves problem (1), works as

follows:

 software client establishes a connection to the

server, having agreed the context. This

determines the rules for creating the most

useful images according to the user's

professional identity. It is believed that the

system has a description of sets of rules for

several contexts. The default context is

specified when the user logs on. The context

does not change in the session, however, with

the geoservice, the same user can set an

arbitrary number of sessions;

 user-analyst sends

qqq ,...,

queries, forming

the skeleton

of the workspace. The

skeleton environment is built on the server

side by applying the

),( BR

knowledge base

rules. The rules implement a reasonable

strategy for building the most useful

workspace. After processing each request, the

server evaluates the utility

)(wI

by changing

the composition and number of environment

objects so as to maximize the utility level;

 on the client side, the analyst uses visualization

tools that are wrappers for standard scaling,

panning, ang and views. Each wrapper uses

the knowledge of the service to display the

required portion of the workspace of the map.

Intelligent selection of environment objects

precedes the standard rendering operation;

 analyst has the ability to disable and re-enable

intelligent geoservice support at any one time.

Disabling means that the application of the

),( BR

rules is blocked. After that, the

analyst continues to study and modify the

workspace manually. The subsequent

inclusion of intellectual support leads to the

reorganization of the workspace: the created

skeleton is provided with a newly constructed

environment;

 session end means either the end of the work

with the geoservice, or the transition to a new

context of visual analysis. In the latter case,

the skeleton of the completed session is saved.

A new session represents the workspace in a

new context by constructing the corresponding

environment.

Analyzing the principle of geoservice, the

following problems should be noted. The quality of

the solution of the problem (1) is determined by the

knowledge of

),( BR

about the objects, classes

and relationships of the



spatial database. If the



structure is modified, i.e. there are previously

unknown classes and instances of objects and

relations, the management of the dialog is losing its

effectiveness for two reasons. First, any new classes

of objects and relations between them are absent as

facts or rules of the knowledge base. Consequently,

their re-use in subsequent sessions is impossible.

Workspaces of analysis lose their quality due to the

lack of relevant data. Secondly, there is an objective

process of "obsolescence" of knowledge about the

dialogue. The classes and instances of objects used

for a certain period of time lose their significance

and are interpreted as redundant. Since geoservices

never delete the accumulated data (Shashi and Hui,

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

438

2008), the work areas are becoming increasingly

redundant. Thirdly, the growth in the number of

objects of known classes makes it increasingly

difficult to select the most useful of them. Additional

knowledge is required for an adequate selection of

the most significant objects.

4 KNOWLEDGE DISCOVERY

PROBLEM DEFINITION

Knowledge in the task of managing of interactive

interaction is understood as information that allows

the geoservice to efficiently select useful data into

the workspace of the analysis. The geoserver in this

case is considered as an evolving system (Angelov,

2012), which accumulates data in real time and self-

learning.

The main source of knowledge is the work area,

which is built by an analyst without intellectual

support of geoservice. What can force a user to

temporarily abandon it? Analysis shows that there

can be three reasons.

The first reason is that geoservice does not use

objects and relationships that are important from the

point of view of the analyst. The information deficit

is replenished by the analyst manually, and this must

be done repeatedly in each session.

The second reason is connected with filling the

work area with insignificant objects, which leads to

an increase in redundancy. This leads to

deterioration in the perception of the cartographic

image, and to an increase in the laboriousness of

manipulating the image in the process of its study.

The third reason is the unsatisfactory work

function of changing the complexity of the

workspace. As mentioned above, this function is an

analog of the known scaling and panning functions

with the only difference that when you execute

them, the composition of the objects of the

workspace changes. This is done to preserve its

logical consistency. The unsatisfactory work of the

function in question is that its integrity is violated.

A manually constructed fragment is a source of

knowledge about the usefulness of a cartographic

image. Obvious is the subjectivity of this

knowledge, its ontological uncertainty and the

vagueness of the evaluation of the actual material.

For this reason, the following approach is proposed:

at the moment when the intellectual support of the

interactive dialogue is disconnected, fix the

manually constructed work area, then, based on the

analysis of the composition of the selected objects,

generate product rules, and then evaluate their

utility. Utility is confirmed by the repeated use of

rules by users of the service.

5 RULES GENERATION

PROCEDURE

The

),( BK

rules display knowledge of how the

work area boundaries (

) are defined for a given

skeleton

and the objects selected for analysis are

ranked by significance. Order on a set of

workspace objects is used to display the most

important elements of the image when zooming and

panning. Accordingly, it is possible the generation of

several types of rules.

The rules for determining the spatial boundary

have the form:

IF Properties (Object) Corresponds Sample

THEN BufferRadius = Value.

Here, the Correspond relation establishes a way

to map the attributes of the object selected by the

Property (x) function to the reference value. The

variable BufferRadius is used to build a spatial

buffer for a given object. The construction of the

buffer zone is a standard analytical function of

geoinformation systems (Shashi and Hui, 2008),

which consists in constructing the convex hull of a

set of geometric primitives at a given distance (of a

given radius). The buffer zones of individual objects

are combined into a single buffer zone of the

skeleton

. An example of such a rule may be:

IF TypeOfObject (ObjectID) = “Road” THEN

BufferRadius = 40,

where TypeOfObject (x) function determines the

type of the object by its identifier.

The rules for determining the temporary border

are presented in a similar way:

IF Properties (Object) Corresponds Sample

THEN Interval = Value.

Here, the variable Interval indicates a deviation

relative to the set base point of time. This point, like

the spatial coordinates of the analyzed area on the

map, can change in the course of work.

The rules for defining semantic boundaries are of

the form:

Knowledge Discovery for Interactive Dialogue Management with Geoinformation Service

439

IF Properties (Object) Corresponds Sample

THEN AddToList (BorderElement).

The AddToList(x) function adds instances of

objects and relationships to the list that forms the

semantic boundary of the workspace.

A spatio-temporal query to the geoservice

database can be constructed from the result of

applying the rules to skeleton

at any one time.

The result of which is the addition to the skeleton of

the environment

Rules of preference on the set of objects of the

workspace have the form:

IF Properties (Object1) Corresponds Sample1

AND Properties (Object2) Corresponds Sample2

THEN Object1 ComparisonOperator Object2.

By the designation ComparisonOperator is

meant a possible variant of comparison - "more

preferable", "less preferred", etc.

The above rules are used to describe the

precedents of constructing the shell of the

workspace. To evolve knowledge from individual

facts to generalizations, rules of another type must

be generated. The structure of these rules is

determined by the logic of using cartographic

information. Consider the following operations for

generating rules: aggregation, generalization and

transposition.

An aggregation of rules will be called an

operation that leads to a new rule from several

existing rules, which involves performing in some

way all the actions of the original rules:

),,...,,(

21 kа

RRRAR 

where A is the aggregation operator. An example

of aggregation may be the disjunction of antecedents

and the conjunction of the consequent of the original

rules. The aggregating rule

is entered in the

knowledge base, and rules

RRR ,...,,

are deleted.

The aggregation result is not reliable, but it looks

plausible. A prerequisite for constructing an

aggregating rule is the following:

,1,







kji

where

is the workspace in which

rule

was previously built. In accordance with this

condition, the aggregating rule

can be generated

only for space-time and semantic areas that have

"something in common".

The generalization of

RRR ,...,,

rule set

involves the construction of a general rule

),,...,

(

RRRGRg 

where G is the generalization operator. As in the

previous case, only the

rule remains in the

knowledge base. Generalization also does not

provide reliable knowledge, although it may turn out

to be plausible. Unlike aggregation, generalization

implies the existence of a correspondence between

the objects of electronic maps of different levels of

generalization. Thus, availability of ready maps of

various levels of generalization is mandatory.

The transfer of experience between different

work areas (transposition of rules) is based on a

logical analogy. The operation of constructing

rule similar to the previously generated

rule

consists in applying the similarity operator

).(

RLR 

This operator implements the principles of

geographical classification of territories. An

example is the proximity of the workspaces by

geographical indicator - the relief or geophysical

structure. The transfer of experience consists in the

transformation of antecedent objects and consequent

of the chosen rule into objects of another workspace.

The operation of transferring experience is

unreliable, but plausible.

6 CONFIRMATION OF RULES

The generation methods considered above give only

hypotheses that require confirmation of their

usefulness. The task of selecting useful rules is to

evaluate the reliability of the generated rules and

exclude unreliable (useless) rules from the

knowledge base of the service. The criterion of

reliability is the degree of collective recognition of

the usefulness of the rule by users of the geoservice

for a limited time. The time factor is necessary for

the following reasons:

 time imitates the "aging" of knowledge. There

is aging information on how to use spatial

data, in contrast to spatial data itself. The

objects and relationships used to solve the

same problem change for many reasons over

time;

 time pragmatically limits the usefulness of the

rules to "the present time". Geoservice does

not set out to develop fundamental rules, with

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

440

a value independent of time. Priority is given

to locality and efficiency of knowledge;

 cartographic production regulates the regulatory

periods for updating maps, schemes and plans.

There is no such requirement with respect to

the knowledge base. This means that

periodically there is an objective need to

develop new knowledge, which should ensure

the adaptation of the service to the updated

cartographic basis.

On this basis, pair

*)*,( nT

, in which

is the

duration of the confirmation time interval, and

the number of confirmations, should be considered a

measure of the confidence of the user service team

in the usefulness of the rules. The rule is considered

to be confirmed if, within the time

*Tt 

from the

moment of the rule's appearance,

*nn 

precedents

of its successful use are fixed. If at least one of the

formulated conditions is violated, the rule is

removed from the geoservice knowledge base.

Analyzing the proposed measure, it is necessary

to note the following:

1) there is a high probability that collectively a

useless rule will be confirmed for small

values.

This is explained both by the psychology of user

behaviour (the ability to tolerate "inconvenience")

and the content of the workspace (in the area under

investigation, the rules may appear weak). The

probability under consideration decreases with

increasing

, and for large values, confirmation of

useless rules becomes impossible;

2) small values

lead to discarding useful

rules. The reason for this may be insufficient

intensity of geoservice use as a whole, or low

activity of analysts' work with separate areas.

Increasing the value of

reduces the probability

of losing useful knowledge;

3) The limiting factor for values

and

values is the server performance. To identify and

verify knowledge, a fixed proportion of productivity

is assigned, which can be represented as the speed of

confirmation of rules

. The values of the pair

*)*,( nT

cannot be chosen uniquely

because

*/* TnV 

It should be noted that speed of information

distribution in the social network can be used as a

limit to the rate of confirmation of the rules. At

present, there are a number of works devoted to

mathematical models of information dissemination

and the influence of users of social networks on each

other (Isea and Mayo-García, 2015). Analysts-users

of geoservice in many cases form professionally

oriented communities that have all the attributes of a

social network, which allows obtaining numerical

values of the rate of distribution of rules.

Note also that the number of

confirmations

accumulates as a result of the summation of not

necessarily integers. As mentioned above, the

rule is confirmed if it is used to form the

environment (

) of the current

work area and

the user-analyst does not disable intelligent

geoservice support. Since each rule has a spatio-

temporal and semantic binding of

obtained at

the time of generation, the confirmation should be

estimated by the value:

.||/|| wwwn





Here,

|| x

denotes the power of a set of

cartographic objects in the region

If the rule is not confirmed, the total number of

confirmations is subtracted from the value

||/||

wwwn 



The above relations take into account the fact

that the scope of each generated rule is limited.

7 CONCLUSION

The considered mechanism of using spatial data by

geoservice gives an effect due to the spread of the

experience of constructing workspaces for visual

analysis. The mechanism is based on knowledge of

the utility of individual objects and relationships

subjectively identified by users. Due to this, the

quality of information content of workspaces is

increasing and the quality of decisions taken on the

basis of spatial data is improved.

The advantage of the described principle is the

exclusion of the rigid dependence of the operation of

the geoservice software components on the structure

of the spatial database. Traditionally, all the

possibilities of forming workspaces are determined

by the database schema. To use new information

objects, you need to publish their description and

modify the software components. The principles of

using knowledge about new information objects

considered in this paper are not limited to a specific

database schema.

The experience of users, fixed by the rules, is of

great importance in the process of geoservice

adaptation to changing the structure of the

information environment. Consequently, the quality

of geoservice work depends heavily on the activity

Knowledge Discovery for Interactive Dialogue Management with Geoinformation Service

441

of users in the search and use of data. In this paper,

we propose the operation of generating hypotheses

based on existing rules to reduce this dependence.

Even with a small experimentally confirmed

material, many hypotheses can be constructed,

which will then be tested by experiment.

Evaluating the effect of the proposed approach,

we should pay attention to reducing the collective

costs of finding useful information. It is known that

the complexity of searching among N objects at best

requires O(logN) operations. Such costs in the case

of individual work are inevitable and are summed up

for all users of the geoservice. When you reuse an

element already found, the complexity is O(1),

which for the user community gives a significant

gain.

Further research can focus on the generalization

and transfer of the evolutionary principle to network

services of another purpose.

ACKNOWLEDGEMENTS

This work has been supported by the Ministry of

Education and Science of the Russian Federation

under Project 2.918.2017.

REFERENCES

Ali, A.L., Schmid, F., 2014 Data Quality Assurance for

Volunteered Geographic Information. In Geographic

Information Science Lecture Notes in Computer

Science, 8728, pp. 126-141.

Andrienko, G., Andrienko, N., Bak, P., Keim, D., Wrobel,

S., 2013. Visual Analytics of Movement. Springer-

Verlag. Berlin.

Andrienko, G., Andrienko, N., Rinzivillo, S., 2016.

Leveraging Spatial Abstraction in Traffic Analysis and

Forecasting with Visual Analytics. Information

Systems,57, pp. 172–194.

Angelov, D., Filev P., Kasabov, N., 2010. Evolving

Intelligent Systems: Methodology and Applications.

John Wiley & Sons.

Angelov, P., 2012 Autonomous Learning Systems: From

Data Streams to Knowledge in Real-time. John Wiley

& Sons.

Belyakov, S. L., Bozhenyuk, A. V., Belykova, M. L.,

Rozenberg, I. N., 2014. Model Of Intellectual

Visualization Of Geoinformation Service. In

Proceedings 28th European Conference on Modelling

and Simulation ECMS 2014, pp.326-332.

Brimicombe, A., Chao L., 2009. Location-Based Services

and Geo-Information Engineering. John Wiley &

Sons.

Faby, H., Koch, A., 2010. From maps to neo-cartography.

In 3rd International Conference On Cartography and

GIS. pp. 86-91.

Fairbairn, D., 2015 Mapping Disorder: An Exploratory

Study. Modern Trends in Cartography. In Lecture

Notes in Geoinformation and Cartography, pp. 13-22.

Goodchild, M. F., Li, L., 2012 Assuring the quality of

volunteered geographic information. In Spatial

Statistics, 1, pp.110-120.

Hart, G., Dolbear, C., 2013. Linked Data: A Geographic

Perspective. Taylor & Francis.

Isea, R., Mayo-García, R., 2015 Mathematical analysis of

the spreading of a rumor among different subgroups of

spreaders. In Pure and Applied Mathematics Letters,

2015, pp 50-54.

Keim, D.A., 2002 Information visualization and visual

data mining. In IEEE Transactions on Visualization

and Computer Graphics, 8, pp. 1-8.

Li, W., Linna, L., Goodchild M.F., Anselin, L., 2012. A

geospatial cyberinfrastructure for urban economic

analysis and spatial decision-making. In ISPRS

International Journal of Geo-Information, 2, pp. 413–

431.

Lughofer, E., 2011. Evolving Fuzzy Systems.

Methodologies, Advanced Concepts and Applications.

Springer Science & Business Media.

Malczewski, J., 2004. GIS-based land-use suitability

analysis: a critical overview. In Progress in Planning,

62. pp. 3–65.

Pettit, C., Cartwright W., Bishop, I., Lowell, K., Puller,

D., Duncan, D., 2008. Landscape Analysis and

Visualisation, Spatial Models for Natural Resource

Management and Planning. Springer-Verlag, Berlin.

Popovich, V., Vanurin, S., Kokh, S., Kuzyonny, V., 2011.

Intellectual Geographic Information System for

navigation safety. In IEEE Aerospace and Electronic

Systems Magazine, 26, pp.29-31.

Shashi, S., Hui, X., 2008. Encyclopedia of GIS.

SpringerScience+Buisiness Media, LLC. New York.

Wang, S., 2010. A cyberGIS framework for the synthesis

of cyberinfrastructure, GIS, and spatial analysis. In

Annals of the Association of American Geographers,

100, pp.535-557.

Yang, C., Sun, M., Liu, K., Huang, Q., Li, Z., Gui,

Z., Yunfeng, J., Xia, J., Yu, M., Xu, C., Lostritto, P.,

Zhou N., 2015. Contemporary Computing

Technologies for Processing Big Spatiotemporal Data.

In Space-Time Integration in Geography and

GIScience, pp. 327-351.

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

442