trigram

This is a table of type trigram and their frequencies. Use it to search & browse the list to learn more about your study carrel.

trigram	frequency
the number of	3475
expert systems with	3036
systems with applications	2964
in order to	2914
based on the	2266
a set of	1829
proceedings of the	1731
t h e	1639
the expert system	1592
an expert system	1568
one of the	1478
the use of	1390
in this paper	1368
as well as	1304
international conference on	1287
in terms of	1192
expert system for	1188
in proceedings of	1166
according to the	1095
can be used	1089
of the system	1066
international journal of	1031
o f the	1017
the set of	976
i o n	947
the performance of	901
is based on	878
as shown in	866
a n d	862
with respect to	858
is used to	852
due to the	847
the knowledge base	846
t i o	846
be used to	773
in this case	772
on the other	771
of the proposed	764
part of the	754
there is a	752
a number of	751
it can be	717
the other hand	715
the results of	713
shown in fig	711
the development of	705
x x x	705
e n t	697
ieee transactions on	689
the value of	685
in which the	671
used in the	646
a a a	627
that can be	623
the fact that	597
a t i	583
of the th	580
performance of the	580
value of the	557
is shown in	557
in this section	550
some of the	546
the case of	544
f o r	543
shown in figure	536
fuzzy expert system	524
it is not	522
there is no	516
the process of	513
each of the	506
of the expert	505
the quality of	498
c o m	494
in the following	493
be able to	486
are used to	485
most of the	480
based expert system	479
in the case	476
such as the	473
shown in table	472
to evaluate the	470
based on a	467
related to the	467
p r o	463
i n g	460
as a result	457
results of the	457
can be seen	456
i i i	452
of expert systems	451
in the first	448
c o n	445
by means of	444
is the number	443
x sbref http	434
in this study	430
the problem of	429
c c c	427
is defined as	427
an example of	415
all rights reserved	413
expert system is	413
to determine the	408
the context of	408
in the same	404
the degree of	402
of the most	401
n n n	398
the system is	396
which can be	396
in the literature	395
depending on the	394
in addition to	393
o n s	391
the size of	388
the application of	387
the accuracy of	383
to find the	382
respect to the	379
of the data	378
m a n	375
h e r	371
to improve the	370
the basis of	369
of this paper	369
all of the	369
total number of	368
the proposed method	368
values of the	368
are shown in	367
the concept of	366
the form of	363
the values of	362
it is possible	361
the amount of	361
this paper is	358
the worker was	357
the inference engine	354
of the problem	351
size of the	350
well as the	347
a t e	347
in this work	347
it is a	343
the selection of	340
can be found	339
use of the	339
is one of	337
m e n	336
is that the	336
large number of	335
the impact of	333
at the same	331
to the user	330
on the basis	329
of an expert	329
need to be	329
a g e	328
analysis of the	327
in the form	326
in the next	322
of the model	321
in the field	319
the level of	317
of computer science	317
to deal with	317
t h a	316
th international conference	313
to be a	313
the result of	311
n o t	310
the importance of	310
by the user	310
in other words	308
in the context	308
it has been	308
the rest of	308
to reduce the	307
that it is	306
information about the	304
p e r	303
the presence of	301
o f a	300
in the second	299
the effect of	297
of expert system	297
the analysis of	295
can also be	294
accuracy of the	294
this is a	294
the total number	294
the type of	294
to obtain the	292
associated with the	291
is able to	291
and it is	291
state of the	290
because of the	290
a variety of	288
t r a	288
are based on	288
a c t	287
it is important	286
user overlap levels	286
the purpose of	286
in this way	285
the structure of	284
the field of	283
that there is	282
on artificial intelligence	281
of the knowledge	281
h a t	281
similar to the	281
to identify the	278
version of the	278
depends on the	277
be found in	277
provided by the	277
european journal of	277
com locate eswa	276
note that the	275
e r e	275
of the two	275
the same time	274
quality of the	273
by using the	273
in the previous	273
of artificial intelligence	272
included in the	270
to be used	270
a list of	270
be used in	269
the training set	269
is possible to	269
of all the	268
the ability to	267
description of the	265
the probability of	265
has to be	265
this is the	265
is important to	264
of the first	263
structure of the	262
figure shows the	261
in the system	261
used as a	260
be used for	260
more than one	260
p p p	259
which is a	259
rule rule rule	258
the th international	257
the end of	257
t e r	257
for a given	257
by the system	256
have to be	256
point of view	256
and so on	255
s t r	255
in computer science	255
is not a	255
expert systems for	255
terms of the	255
i n f	255
of the same	254
there are two	254
e e e	253
h e n	253
a large number	253
in the knowledge	252
one or more	251
expert system to	250
used for the	249
n c e	249
to solve the	248
a series of	248
a total of	248
n f o	247
the most important	247
it should be	246
expert systems in	245
the nature of	245
lecture notes in	244
evaluation of the	244
the proposed approach	243
the role of	242
of operational research	241
the effectiveness of	241
t i c	241
h h h	240
so that the	240
journal of operational	238
in artificial intelligence	238
applied to the	238
the design of	237
a and b	237
be used as	237
m a t	237
based expert systems	236
added to the	235
which is the	235
ti o n	234
o m e	232
described in section	232
the university of	231
in expert systems	230
of user overlap	229
a knowledge base	229
conference on artificial	229
corresponding to the	229
is given by	229
for each of	228
n a l	228
table shows the	228
used by the	228
of the algorithm	228
into account the	227
show that the	227
o r m	227
the length of	227
the user to	227
of the user	226
different types of	226
the definition of	226
w w w	226
the instant case	226
computer science and	225
the sum of	225
the relationship between	224
needs to be	224
of the rule	224
nature of the	224
system for the	223
save alert research	223
alert research feed	223
this means that	222
p a g	222
journal of the	221
the combination of	220
is necessary to	219
a r t	219
system based on	219
with the same	218
e t h	216
corresponds to the	216
ieee international conference	216
given by the	216
may not be	216
the evaluation of	215
to be the	215
o m p	215
o t h	214
t e d	214
compared to the	213
development of a	213
it is necessary	213
to calculate the	212
decision support systems	212
e r a	211
the complexity of	211
it is also	210
department of computer	210
was used to	210
taking into account	210
r a c	210
function of the	209
c a t	209
are presented in	208
used in this	207
there are several	207
we use the	206
have been proposed	206
r e s	206
to estimate the	206
obtained from the	206
that the proposed	205
of the paper	205
o u r	205
is organized as	204
it does not	204
the choice of	203
is used for	203
the state of	202
is presented in	202
a n u	202
the user is	202
the aim of	201
and the number	201
referred to as	201
expert systems and	201
the rule base	200
d e f	200
of the following	199
implementation of the	199
science and technology	199
to use the	199
rest of the	198
can be obtained	198
the percentage of	198
s t e	198
n t e	198
length of the	197
can be applied	197
we propose a	197
is used as	196
organized as follows	196
and can be	195
f f e	195
t e m	194
at the end	194
i n t	194
the proposed system	194
representation of the	194
of the number	194
to test the	194
it is the	193
refers to the	193
a p p	193
in the data	192
decision support system	192
b l e	192
v e r	192
knowledge of the	192
this can be	192
means that the	192
of a set	192
it is definite	191
if there is	191
analytic hierarchy process	191
we can see	191
the data set	190
n o w	190
the difference between	190
features of the	190
have been used	190
in such a	190
a combination of	189
to represent the	189
is used in	189
the results are	189
c t i	189
is the most	189
the cost of	189
to select the	189
o n e	188
a case study	188
worker was not	188
to develop a	188
in this research	188
changes in the	187
according to their	187
that of the	187
t t t	186
f e f	186
e i n	186
there is an	186
a r e	186
in the training	186
user overlap level	186
f f f	186
the implementation of	185
support vector machines	185
see front matter	185
the range of	183
r a t	182
at least one	182
e x p	182
of the input	182
artificial intelligence and	182
better than the	182
n is the	182
x and y	181
the next section	181
to measure the	181
stored in the	181
the proposed algorithm	181
a fuzzy expert	181
any of the	180
consists of a	180
d d d	180
in the last	180
fuzzy expert systems	179
this paper we	179
f e e	179
a i n	179
have the same	179
de db conf	179
the user interface	179
a subset of	179
in case of	178
the goal of	178
has been used	178
of the other	178
application of the	178
d m a	177
by the expert	177
have been developed	177
in the process	176
difference between the	176
this type of	176
is given in	176
to find a	176
result of the	176
be applied to	176
it may be	176
and data mining	175
expert system in	175
is applied to	175
paper is organized	175
taken into account	175
e v e	175
e m e	175
to have a	174
the system was	174
e r t	174
e f f	174
the diagnosis of	174
that have been	174
the first step	174
the efficiency of	173
r m a	173
is to be	173
e f e	173
can be represented	173
and in the	172
use of a	172
in the future	172
set of rules	172
characteristics of the	172
as it is	172
used as the	171
is less than	171
of the process	171
defined as the	171
it would be	171
k n o	171
compared with the	171
they do not	171
the results obtained	171
of the three	171
to provide a	171
to make the	170
to the system	170
r t i	170
a result of	170
the training data	170
a b s	170
shown in the	170
a collection of	169
the task of	169
the possibility of	169
the majority of	169
take into account	169
the list of	168
is composed of	168
complexity of the	168
of the rules	168
the system to	168
b s t	168
results for the	168
development of the	168
we do not	168
fuzzy sets and	168
a n t	168
we need to	167
is as follows	167
o o o	167
the user can	167
to assess the	167
this kind of	167
of the results	167
the characteristics of	166
the output of	166
the existence of	166
of the ieee	166
as an example	166
c l e	165
the effects of	165
o r e	165
of machine learning	165
org licenses by	164
for all the	164
is the set	164
o n t	164
of the form	164
so as to	164
the scope of	164
example of a	163
accepted manuscript a	163
m b e	163
to create a	163
expert systems are	163
each of these	163
the construction of	163
the paper is	162
the distribution of	162
hindawi publishing corporation	162
results show that	162
to compute the	162
artificial neural networks	162
will be used	162
fact that the	162
between the two	161
is equal to	161
support vector machine	161
o n a	161
to the fact	161
the same as	160
journal of advanced	160
likely to be	160
the employer paid	160
the need for	160
to predict the	160
impact on the	160
as described in	160
can be defined	160
and the other	160
d e s	160
b d e	159
of the acm	159
a measure of	159
as in the	159
to the number	159
to obtain a	159
system can be	158
be considered as	158
of the main	158
in which a	158
shows that the	158
parts of the	158
found in the	158
t c t	158
there are many	158
the objective function	157
o u t	157
p a r	156
number of features	156
information from the	156
t h o	156
to be able	156
is a set	156
i t i	156
is represented by	155
b e r	155
if it is	154
e e f	154
publishing corporation http	154
the behavior of	154
the creation of	154
as part of	154
described in the	154
addition to the	154
of a rule	153
end of the	153
of the current	153
set o f	153
generated by the	153
can be considered	153
the lack of	153
would like to	152
is related to	152
e s t	152
the domain of	152
in the current	152
the nearest neighbour	152
in recent years	152
the ability of	152
expert system and	152
the identification of	151
model of the	151
as the number	151
are able to	151
of the fuzzy	151
article in press	151
equal to the	151
notes in computer	151
the similarity between	151
m o r	151
a sequence of	151
in the domain	151
can be easily	150
y s t	150
that are not	150
the distance between	150
to the same	150
that the system	150
two types of	150
this is an	150
is similar to	150
in centimeters is	149
presented in the	149
in the past	149
of the best	149
there are no	148
l e i	148
the weight of	148
a m e	148
on machine learning	148
a way that	147
a t a	147
s y s	147
of the domain	147
y y y	147
of the new	147
defined by the	147
e a s	146
in the development	146
view of the	146
a ti o	146
they can be	146
of the information	145
comparison of the	145
m o d	145
of a given	144
e n c	144
i n e	144
in contrast to	144
a group of	144
the objective of	144
m e t	144
are given in	144
this is because	144
the area of	144
of the decision	144
are used in	144
e p a	144
is defined by	144
has been developed	143
the sth percentile	143
a neural network	143
is difficult to	143
fuzzy set theory	143
x p e	143
r e n	143
of the original	142
and only if	142
a framework for	142
parameters of the	142
of the work	142
overview of the	142
values for the	142
the meaning of	142
i c l	142
n u m	142
the generation of	141
is set to	141
the notion of	141
s s s	141
f b d	141
r o m	141
understanding of the	141
v v v	141
be noted that	141
t i n	141
of the time	140
supported by the	140
aspects of the	140
a function of	140
a lot of	140
to increase the	140
is an important	139
john wiley sons	139
is associated with	139
with the highest	139
based on their	139
the problem is	139
number of rules	139
and expert systems	139
to describe the	139
a s e	139
context of the	138
is described in	138
c c e	138
in the proposed	138
o r t	137
the model is	137
i c a	137
in the database	137
is greater than	137
e r s	137
out of the	137
e d e	137
in the rule	137
o w l	137
u m b	137
case of the	137
school of business	137
of the test	136
artificial neural network	136
e s e	136
involved in the	136
to the best	136
are used for	136
number of the	136
upper and lower	136
by using a	136
there are some	135
t a t	135
presented in this	135
the order of	135
research and development	135
expert system shell	135
the analytic hierarchy	135
to model the	135
that there are	135
a model of	134
van den poel	134
the idea of	134
the semantic web	134
the results for	133
for this reason	133
in our case	133
associated with a	133
an overview of	133
of the international	133
d a t	132
e c t	132
the expert systems	132
extracted from the	132
level of the	132
expert system based	132
db conf semweb	132
represented by a	132
expert system was	132
of the art	132
f r o	132
conf semweb om	132
we want to	132
the knowledge of	132
of this study	132
m i n	132
n d i	131
is easy to	131
contents lists available	131
to compare the	131
the time of	131
o n o	131
is the same	131
lists available at	131
m e a	131
p r e	131
they are not	131
m p o	131
joint conference on	131
the system and	131
yes yes yes	130
h o m	130
properties of the	130
selection of the	130
in some cases	130
natural language processing	130
presented in section	130
the parameters of	130
n t i	130
considered to be	130
section describes the	130
example of the	129
a t t	129
e l s	129
because it is	129
if and only	129
data mining and	129
l o c	129
deal with the	129
this is not	129
the search space	129
be seen in	129
to the problem	128
for this purpose	128
derived from the	128
we used the	128
for expert systems	128
the neural network	128
e s s	128
a part of	128
journal of computer	128
it is clear	128
is that it	128
is due to	128
approach to the	127
for the first	127
the theory of	127
do not have	127
r n a	127
focus on the	127
present in the	127
whether or not	126
sum of the	126
is designed to	126
goal is to	126
interpretation of the	126
o c a	126
it will be	126
e s w	126
a decision of	126
m a y	126
of the total	126
the system can	126
in the design	126
output of the	126
in the model	126
wide range of	126
sciencedirect expert systems	126
it is difficult	125
to make a	125
m a r	125
the th percentile	125
could be used	125
found to be	124
a cc ep	124
e c o	124
published by elsevier	124
te d m	124
presented in table	124
e v i	124
expert system can	124
k k k	124
to ensure that	124
neural networks and	124
is equivalent to	124
manuscript a cc	124
ep te d	124
cc ep te	124
in the area	124
sc ri pt	124
world wide web	124
l e m	124
the basis for	124
n u sc	124
in relation to	124
system is a	124
in the sense	124
u sc ri	124
review of the	124
in comparison to	123
set of all	123
the experimental results	123
rules in the	123
i e r	123
t e e	123
relationship between the	123
to produce a	123
for a particular	122
knowledge base and	122
positive and negative	122
are related to	122
determined by the	122
number of clusters	122
the high court	122
this approach is	122
s e v	122
r a m	122
x is a	122
a n a	121
the department of	121
u r n	121
correspond to the	121
the next step	121
maximum number of	121
the construction industry	121
m m m	121
with regard to	121
is in the	120
allows us to	120
depend on the	120
the representation of	120
a human expert	120
the proposed model	120
is not known	120
for the same	120
to that of	120
the study of	120
l s e	119
of the cases	119
the influence of	119
was to be	119
of the second	119
of the training	119
v i e	119
to each other	119
of this research	119
that has been	119
of the target	119
input and output	119
a range of	119
a and a	119
of the network	119
r e a	119
be one of	118
b e e	118
is not possible	118
knowledge base is	118
to expert systems	118
t o r	118
on the same	118
effectiveness of the	118
s t a	118
should be noted	118
t a n	117
n o f	117
te m s	117
the frequency of	117
be seen as	117
paid to the	117
cited on page	117
d e d	117
j o u	117
the knowledge engineer	117
the absence of	117
t o f	116
international workshop on	116
to generate a	116
as can be	116
small number of	116
the classification of	116
the method of	116
distribution of the	116
o m a	116
this work is	116
access to the	116
u l e	116
ys te m	115
of the applicant	115
n o n	115
to support the	115
of the high	115
h o w	115
e e s	115
we have to	115
to build a	115
s ys te	115
o u n	115
conference on machine	115
the detection of	115
in this area	115
advantage of the	115
were used to	115
m e p	115
that do not	115
development of an	115
o n d	114
a l h	114
of the image	114
zhang et al	114
r o f	114
f a c	114
science and applications	114
been used to	114
different user overlap	114
s w a	114
g g g	114
cite this article	114
precision and recall	114
it is worth	114
is capable of	114
knowledge about the	114
method based on	114
and future work	114
l h o	113
can be made	113
system has been	113
the success of	113
such that the	113
p o r	113
are to be	113
in accordance with	113
a review of	113
and the results	113
e o f	113
ieee computer society	113
system and the	113
the advantages of	113
a wide range	112
a b c	112
this section we	112
have not been	112
i t h	112
m l o	112
close to the	112
the beginning of	112
seems to be	112
from the user	112
the best solution	112
to minimize the	112
order to obtain	112
f d f	112
o r a	112
contained in the	112
the most common	112
o m l	112
assigned to the	112
s e n	112
for computational linguistics	112
obtained by the	112
e a c	111
a value of	111
many of the	111
the test set	111
n a n	111
indicates that the	111
belong to the	111
the system will	111
in a particular	111
the work of	111
affected by the	111
been used in	111
is divided into	111
time of the	111
l e d	111
com science journal	110
association for computational	110
of the object	110
and number of	110
details of the	110
b b b	110
can be described	110
elements of the	110
c e d	110
artificial intelligence in	110
a survey of	110
of intelligent systems	110
will not be	110
c t c	110
the same way	110
basis of the	110
the need to	110
in the construction	109
can be calculated	109
i v e	109
n n h	109
h n n	109
f d e	109
o d u	109
be viewed as	109
of fuzzy logic	109
is clear that	109
what is the	109
is referred to	109
pattern analysis and	109
should not be	109
assume that the	109
n t s	109
to be done	109
can be expressed	109
science journal http	108
rules can be	108
in this example	108
we assume that	108
a comparison of	108
the sequence of	108
than that of	108
to be more	108
the elements of	108
l e s	108
of the different	108
is called the	108
u r e	108
considered as a	108
the help of	108
the strength of	108
a c c	108
model for the	108
be defined as	108
of the project	108
increase in the	108
but it is	108
of the solution	108
behavior of the	107
the user overlap	107
results in the	107
with applications j	107
x i j	107
to the other	107
p u t	107
e n d	107
is to find	107
applications j o	107
less than the	107
mapped to a	107
along with the	107
of a new	107
the work was	107
not known whether	107
model based on	107
e r o	107
her his own	106
the most popular	106
greater than the	106
ltd expert systems	106
advanced computer science	106
in most cases	106
the rules of	106
u n d	106
in the experiments	106
sets and systems	106
a n y	106
to the knowledge	106
of advanced computer	106
t e n	106
focused on the	106
in a given	106
that we have	106
to provide the	106
systems have been	106
the position of	106
p l e	105
a n c	105
o m m	105
publishing ltd expert	105
e m a	105
differences between the	105
the validity of	105
m p l	105
average number of	105
does not have	105
is assigned to	105
some of these	105
the features of	105
journal of engineering	105
is not the	105
data in the	105
this paper presents	105
such a way	105
university of technology	105
has not been	105
of the research	105
performed by the	105
can be observed	105
procedia computer science	104
represented by the	104
to the following	104
available in the	104
the occurrence of	104
by the following	104
n t r	104
area of the	104
x h http	104
defined as follows	104
e m p	104
would be one	104
d f b	104
that will be	104
p o s	104
aspect of the	104
in this context	104
analysis and machine	104
n d e	104
e f d	104
there are three	104
the management of	104
position of the	104
of this work	103
we use a	103
d g e	103
r a l	103
o d e	103
e e d	103
carried out by	103
can be done	103
please cite this	103
the chattel was	103
n i n	103
the algorithm is	103
discussed in section	103
that could be	103
and machine intelligence	103
together with the	103
a c h	103
a new approach	103
majority of the	103
the proportion of	102
l l l	102
systems in the	102
r o u	102
knowledge discovery and	102
o n c	102
the knowledge acquisition	102
e d g	102
d e e	102
e a t	102
the real world	102
information on the	102
to analyze the	102
international joint conference	102
h a n	102
i m p	102
based on this	102
assumed to be	102
model can be	102
input to the	102
expert system that	102
b c b	102
e d f	102
to achieve the	102
importance of the	102
means of a	102
to achieve a	101
objective is to	101
is the case	101
solve the problem	101
the integration of	101
particle swarm optimization	101
g e www	101
conference on information	101
that the user	101
is a very	101
defined as a	101
h o d	101
the benefits of	101
the evolution of	101
choose and book	101
o u s	101
we believe that	101
the journal of	101
was based on	100
transactions on systems	100
are summarized in	100
can then be	100
the best results	100
the weights of	100
was carried out	100
two or more	100
the calculation of	100
to avoid the	100
and for the	100
can see that	100
given in table	100
and the worker	100
consists of the	100
to the next	100
known whether the	100
the other two	100
due to its	100
decision of the	100
represented as a	100
the name of	99
a b o	99
knowledge in the	99
at the beginning	99
e e n	99
when it is	99
number of iterations	99
the ratio of	99
of the american	99
are described in	99
to demonstrate the	99
of the search	99
is determined by	99
may lead to	99
are as follows	99
to show the	99
to get the	99
expert systems have	99
available at sciencedirect	99
level in the	99
the target domain	99
n h h	99
the comparison of	99
illustrated in figure	99
m o n	98
conference on computer	98
n t o	98
rule based expert	98
is an open	98
or equal to	98
c h e	98
o w n	98
in our experiments	98
the best of	98
in the original	98
c o u	98
of neural networks	98
be regarded as	98
efficiency of the	98
blackwell publishing ltd	98
n t a	98
estimation of the	97
performed on the	97
has been shown	97
proposed in this	97
l a t	97
on the user	97
c a l	97
section presents the	97
the previous section	97
initial cluster centers	97
for the purpose	97
impact of the	97
the employer and	97
at sciencedirect expert	97
with the help	97
employer and the	97
solution to the	97
to carry out	97
the architecture of	96
components of the	96
intention that the	96
of a particular	96
a pair of	96
for which the	96
with each other	96
a decision tree	96
n h n	96
based on fuzzy	96
be used by	96
are the most	96
of the method	96
as a consequence	96
the advantage of	96
if there are	96
it possible to	96
w l e	96
the sense that	96
one of these	95
the data and	95
to understand the	95
to generate the	95
in the paper	95
d e r	95
the prediction of	95
period of time	95
t i m	95
combination of the	95
design of the	95
h e m	95
a method for	95
the input data	95
court of australia	95
the most relevant	95
i s t	95
d e n	95
and the system	95
at the university	95
the user in	95
cases in which	95
in natural language	95
are in the	94
structural light framing	94
algorithms by varying	94
can be performed	94
as it can	94
in the real	94
in that case	94
be represented as	94
number of times	94
that is not	94
the result is	94
of the world	94
on pattern analysis	94
h e s	94
values in the	94
system in the	94
the addition of	94
s h o	94
change in the	94
to a given	94
it is very	94
of a single	94
measure of the	94
used to evaluate	94
for the user	94
wu et al	94
the international conference	94
architecture of the	94
in the range	94
aim is to	94
in the present	94
is to use	93
the time series	93
to which the	93
new south wales	93
to define the	93
of data mining	93
the first one	93
to the worker	93
found that the	93
v a l	93
o v e	93
as opposed to	93
has been proposed	93
f f d	93
as a function	93
be divided into	93
the details of	93
of the above	93
m e d	93
for the development	93
e t e	93
a d e	93
make use of	93
params is empty	93
knowledge and data	93
wang et al	93
leads to a	93
effect of the	93
a data set	93
is required to	93
d u c	92
the relationships between	92
owner of the	92
definition of the	92
the part of	92
of the four	92
tion of the	92
is the total	92
the content of	92
of a specific	92
a description of	92
of the feature	92
it is easy	92
we present the	92
e p r	92
discovery and data	92
are used as	92
the user and	92
part of a	92
proposed expert system	92
employer paid to	92
and the second	92
the money that	92
e a n	92
legal expert system	92
the description of	92
that the employer	92
in this article	92
stated to be	92
money that the	92
that it can	92
a l l	92
e c e	91
there are a	91
the other claimant	91
chen et al	91
in the table	91
in the test	91
used to represent	91
used to determine	91
we consider the	91
an important role	91
are likely to	91
on the contrary	91
the optimal solution	91
the membership function	91
an introduction to	91
in the number	91
to investigate the	91
belonging to the	91
can be achieved	91
divided into two	91
this method is	90
than the other	90
for digital economy	90
at a time	90
to this end	90
we describe the	90
s o f	90
which has been	90
large amount of	90
this process is	90
the interpretation of	90
fuzzy inference system	90
results from the	90
we present a	90
of our approach	90
at the time	90
is called a	90
digital economy research	90
value of a	90
of the selected	90
can be viewed	90
center for digital	90
p o n	90
the introduction of	89
this paper proposes	89
and the user	89
research stem school	89
if the user	89
knowledge base of	89
of the features	89
represented in the	89
of the ontology	89
to perform the	89
m a i	89
to control the	89
economy research stem	89
the th annual	89
stem school of	89
rough set theory	89
to determine whether	89
as illustrated in	89
subset of the	89
the goal is	89
journal of machine	89
an independent contractor	89
on neural networks	89
made by the	89
that must be	89
i n i	89
r o b	89
expert system with	89
r e p	89
the search for	89
w i t	89
for the diagnosis	89
e m o	89
of science and	89
and data engineering	88
carried out in	88
is carried out	88
it could be	88
w h e	88
the case study	88
can lead to	88
to the expert	88
r r r	88
of the application	88
d m j	88
u c t	88
assessment of the	88
is illustrated in	88
in the expert	88
summary of the	88
in a way	88
is also a	88
is proposed to	88
to use a	88
by applying the	88
to be considered	88
is obtained by	88
form of a	88
the results show	88
the behaviour of	88
r u l	88
then it is	88
neural networks for	88
are stored in	87
presented to the	87
in information retrieval	87
been applied to	87
a genetic algorithm	87
the rules are	87
are included in	87
c f e	87
that is to	87
r a n	87
the system has	87
been proposed to	87
study of the	87
r t a	87
weight of the	87
the decision maker	87
of the web	87
e d i	87
and the corresponding	87
o f t	87
the legal expert	87
results in a	87
this expert system	87
distance between the	86
different from the	86
e c c	86
is better than	86
the true owner	86
algorithm for the	86
that they are	86
d e c	86
evaluate the performance	86
could not be	86
expert system development	86
of th e	86
journal of intelligent	86
used in a	86
in a single	86
due to a	86
machine learning techniques	86
because they are	86
information of the	86
e b e	86
expert system approach	86
c b c	86
consists of two	86
indicate that the	86
can be classified	85
to build the	85
be seen from	85
group decision making	85
that should be	85
the united states	85
cambridge university press	85
in the ontology	85
approaches have been	85
of the patient	85
be carried out	85
it is shown	85
we have used	85
standard deviation of	85
that are used	85
corresponds to a	85
appear in the	85
results and discussion	85
also be used	85
the system in	85
is responsible for	85
given in the	85
of the program	85
in the decision	85
when the user	85
where n is	85
in the set	85
which is not	85
r e d	85
data mining techniques	85
is a function	85
was used for	84
to study the	84
the average of	84
of business ivorking	84
to note that	84
a c o	84
defined in the	84
set of the	84
cases in the	84
transactions on pattern	84
which means that	84
illustrated in fig	84
all of these	84
to facilitate the	84
of the algorithms	84
ivorking paper is	84
data from the	84
the following sections	84
listed in table	84
a small number	84
d e l	84
business ivorking paper	84
have also been	84
is considered as	83
extension of the	83
o o d	83
of the inference	83
the decision tree	83
classification of the	83
p p r	83
of the individual	83
as a whole	83
one of them	83
institute of technology	83
methods have been	83
the original data	83
stage of the	83
in the third	83
into the system	83
to be performed	83
an open access	83
in the network	83
of information technology	83
s t i	83
is said to	83
c e e	83
none of the	83
is known as	83
lead to a	83
in all the	83
instead of the	83
to a set	83
case of a	83
has been applied	82
showed that the	82
this study is	82
shown to be	82
as mentioned in	82
means of the	82
used to identify	82
such a system	82
for the selection	82
is very important	82
in the temporal	82
no no no	82
which may be	82
the first stage	82
b a s	82
of a knowledge	82
of the environment	82
the proposed expert	82
with the system	82
height in centimeters	82
d i n	82
employer did not	82
o p e	82
from the same	82
the data in	82
reported in the	82
of the whole	82
supply chain management	82
the employer did	82
e d a	82
be added to	82
and they are	81
by comparing the	81
were used for	81
to the data	81
and machine learning	81
to illustrate the	81
a s s	81
can be a	81
is likely to	81
we found that	81
number of attributes	81
international symposium on	81
refer to the	81
using the same	81
regardless of the	81
of the english	81
this work was	81
known as the	81
machine learning research	81
lee et al	81
on the one	81
in the above	81
paper is to	81
in this experiment	81
v e l	81
judges of the	81
of the previous	81
the second stage	81
a tool for	81
calculation of the	81
summarized in table	81
an extension of	81
the point of	81
and of the	80
of a system	80
the significance of	80
are the same	80
the availability of	80
on knowledge and	80
produced by the	80
to the set	80
to construct the	80
there has been	80
can be identified	80
domain in the	80
to the current	80
r t h	80
members of the	80
component of the	80
and pattern recognition	80
in the problem	80
of a fuzzy	80
we will use	80
on the number	80
manner in which	80
i m a	80
on knowledge discovery	80
h h n	80
used as an	80
c a n	80
information in the	80
of such a	79
f o keywords	79
than or equal	79
i n c	79
presented in figure	79
in the image	79
the contents of	79
in section we	79
that the relationship	79
the most appropriate	79
we can use	79
an analysis of	79
of the corresponding	79
purpose of this	79
e e c	79
is the first	79
a b l	79
said to be	79
not have a	79
and a set	79
as a set	79
centimeters is the	79
take advantage of	79
on the use	79
through the use	79
each of which	79
and target domain	78
of the set	78
e a d	78
different levels of	78
to verify the	78
we would like	78
was able to	78
s u m	78
the one hand	78
to form a	78
a kind of	78
to choose the	78
to a specific	78
for earthquake prediction	78
a fuzzy set	78
then used to	78
is needed to	78
f e c	78
it is based	78
d f e	78
is added to	78
which is used	78
consistent with the	78
it must be	78
the fuzzy expert	78
we focus on	78
g r a	77
out to be	77
a computer program	77
objective of this	77
that all the	77
information and knowledge	77
the probability that	77
number of possible	77
to address the	77
the first two	77
s h i	77
order to improve	77
the manner in	77
is higher than	77
based on an	77
the issue of	77
order to make	77
the inference process	77
the class of	77
the th acm	77
according to a	77
based approach to	77
the most similar	77
can be improved	77
step in the	77
takes into account	77
to train the	77
is expected to	77
o f this	77
the types of	77
to detect the	77
the risk of	77
systems based on	77
of the task	77
on the data	77
is an example	76
the rules in	76
the properties of	76
the user has	76
to develop an	76
and does not	76
the usage of	76
this research is	76
are associated with	76
at this point	76
conference on knowledge	76
the estimation of	76
of each of	76
h i p	76
process of the	76
see that the	76
a t h	76
to enhance the	76
to be made	76
is represented as	76
applications of artificial	76
b b c	76
number of data	76
to be an	76
the fitness function	76
on computer vision	76
of the study	76
with the following	76
may be used	76
different kinds of	76
m is the	76
in real time	76
with the other	76
a basis for	76
an attempt to	76
the design and	76
the remainder of	76
as explained in	76
and their applications	76
is calculated as	76
we have a	76
the reason for	75
the limitations of	75
from the data	75
it is an	75
the likelihood of	75
t o m	75
is calculated by	75
step of the	75
yang et al	75
t a i	75
the maximum number	75
based on these	75
b o u	75
e r i	75
table presents the	75
in the final	75
and that the	75
there may be	75
the terms of	75
in the world	75
the employee specification	75
distance from the	75
operation of the	75
o f rules	75
training and testing	75
which have been	75
on expert systems	75
as long as	75
lower and upper	75
nearest neighbour is	75
the fuzzy set	75
be represented by	75
can be divided	75
to learn the	75
the proposed algorithms	75
that may be	75
validity of the	75
an inference engine	75
with a set	75
t and t	75
this information is	75
way in which	75
the standard deviation	75
such as a	75
expert system has	75
only a few	75
approach based on	75
of using the	75
proposed in the	75
o r i	75
the data sets	74
this case the	74
by varying user	74
the mean of	74
to find out	74
validation of the	74
e p t	74
a m o	74
used to calculate	74
had to be	74
which are not	74
to reach a	74
the course of	74
the applicability of	74
p and p	74
varying user overlap	74
on how to	74
considered in the	74
overlap level in	74
of the nd	74
important role in	74
th e d	74
to focus on	74
is done by	74
when there is	74
divided by the	74
that in the	74
u t e	74
of a user	74
com science article	74
computer vision and	74
dependent on the	74
behaviour of the	74
the domain knowledge	74
o s e	74
and on the	74
into a single	74
the problem domain	74
points in the	74
more or less	74
be taken into	74
conclusions and future	74
if they are	74
c and c	74
be obtained by	74
order to achieve	74
ability of the	74
discussed in the	74
communications of the	74
the higher the	74
high court of	74
problem can be	73
page page page	73
be seen that	73
the direction of	73
has also been	73
by the decision	73
widely used in	73
t u r	73
in press as	73
r o p	73
beginning of the	73
be possible to	73
c e s	73
number of nodes	73
e m s	73
as far as	73
relevant to the	73
probability of the	73
is a good	73
is to provide	73
of the existing	73
s y n	73
to take into	73
open access article	73
the decision making	73
stored in a	73
to do so	73
of the case	73
and expert system	73
evolution of the	73
on the web	73
in the evaluation	73
the concepts of	73
rest of this	73
scheduled start date	73
as discussed in	73
consists of three	73
those of the	73
yes no no	73
target start date	73
for a specific	73
does not require	72
the nearest other	72
each one of	72
accused did not	72
start date scheduled	72
an integral part	72
asset location description	72
which are the	72
shows an example	72
the test data	72
for the system	72
cost of the	72
by the number	72
that the number	72
with applications xxx	72
b c d	72
role in the	72
the proposed framework	72
scheduled end date	72
viewed as a	72
support system for	72
rather than the	72
the problems of	72
this article in	72
the rule is	72
to overcome the	72
d o m	72
description actual end	72
location description actual	72
o b l	72
to classify the	72
c h a	72
higher than the	72
form of the	72
machine learning and	72
with the aim	72
not possible to	72
been used for	72
like to thank	72
s i n	72
half of the	72
date scheduled end	72
be included in	72
date target start	72
i t y	72
it is assumed	72
a way to	72
of legal reasoning	72
the support of	72
in the example	72
in the s	71
amount of information	71
with the user	71
t t e	71
integral part of	71
national conference on	71
the accused did	71
b u t	71
will be the	71
aim of this	71
portion of the	71
the change in	71
c h o	71
with the use	71
more and more	71
comparison between the	71
in a similar	71
and analysis of	71
of which is	71
used in our	71
idea is to	71
of them are	71
was applied to	71
journal of production	71
a study of	71
given to the	71
the relationship would	71
value for the	71
to the first	71
a system for	71
are similar to	71
set of features	71
nodes in the	71
r e l	71
described in this	71
to validate the	71
relationship would be	71
which will be	71
m p t	71
rules of the	71
the speed of	71
in more detail	71
e s p	71
the web page	71
springer berlin heidelberg	71
c e p	70
and knowledge discovery	70
data set is	70
the learning process	70
acm transactions on	70
association rule mining	70
n d a	70
is a need	70
type of the	70
shown that the	70
in the input	70
e a r	70
to the development	70
can be written	70
d i t	70
pattern recognition letters	70
to cope with	70
c lu d	70
the outcome of	70
should be able	70
an increase in	70
as for the	70
made up of	70
to implement the	70
of the key	70
science and engineering	70
of the st	70
of the human	70
lu d e	70
that the expert	70
s o m	70
set of attributes	70
decision did not	70
so that it	70
and information science	70
and its application	70
can only be	70
the average number	70
be described as	70
t o t	70
to consider the	70
since it is	70
to create the	70
is that they	69
the p r	69
w it h	69
used to generate	69
of the chattel	69
of the accused	69
of information and	69
conference of the	69
and the average	69
model and the	69
seen as a	69
in t h	69
is considered to	69
to the original	69
positive or negative	69
the feature extraction	69
r r e	69
at time t	69
each pair of	69
held that the	69
would not be	69
side of the	69
order to evaluate	69
in a specific	69
t r o	69
first of all	69
an application of	69
s e d	69
with the expert	69
wide variety of	69
to perform a	69
we define the	69
back to the	69
on the results	69
t r u	69
ensure that the	69
followed by a	69
with different user	69
also known as	69
all of them	69
to examine the	69
to one of	69
and implementation of	69
presented at the	69
computer and information	69
by the same	69
building expert systems	69
the temporal dimension	69
process can be	69
noted that the	69
of our knowledge	69
many times on	68
can be computed	68
e f c	68
leads to the	68
decision making in	68
same as the	68
and artificial intelligence	68
in the text	68
european conference on	68
average of the	68
algorithm based on	68
t i v	68
and then the	68
on a set	68
for the second	68
of the third	68
the input and	68
has been implemented	68
s o n	68
may also be	68
the assessment of	68
the way in	68
temporal and location	68
the solution of	68
for the three	68
be expressed as	68
employer paid the	68
to reach the	68
for the case	68
relative to the	68
r t s	68
paid the worker	68
techniques have been	68
results of this	68
in a rule	68
in the presence	68
times on page	68
to check the	68
name of the	68
performance of a	68
p a p	68
e n e	68
observed that the	68
in a real	68
some of them	68
c e r	68
which is based	68
as one of	68
order to avoid	68
n a t	68
are applied to	68
of the concept	68
of new south	68
the feasibility of	68
by elsevier ltd	68
method for the	68
a lack of	68
of the rd	68
cited many times	68
knowledge base for	67
applied to a	67
it consists of	67
was supported by	67
weights of the	67
d a n	67
a r a	67
of data points	67
step is to	67
date maintenance operator	67
what are the	67
to a descriptor	67
system as a	67
the similarity of	67
a s u	67
propose a new	67
the decision did	67
n e w	67
c e f	67
from the set	67
of this approach	67
of the tree	67
e l a	67
described as follows	67
a decision support	67
for the analysis	67
u n i	67
purpose of the	67
each of them	67
and information technology	67
choice of the	67
expert systems can	67
has been a	67
from a set	67
t a b	67
legal expert systems	67
for the next	67
the natural specification	67
d e a	67
the computation of	67
the genetic algorithm	67
was set to	67
the top of	67
divided into three	67
the location of	67
of the variables	67
the most recent	67
cnc turning centre	67
amount of data	67
the differences between	67
li li li	67
e d u	67
top of the	67
d c c	67
in a more	67
n s t	67
u g h	67
end date maintenance	67
of the k	67
transactions on information	67
the decision affected	67
transactions on knowledge	66
and decision making	66
the most frequent	66
the c o	66
values for each	66
p t i	66
three types of	66
in the user	66
division of the	66
there was no	66
m e m	66
and neural networks	66
from the training	66
best of our	66
of these two	66
be difficult to	66
c c b	66
the information from	66
using the following	66
of the evaluation	66
of the machine	66
to develop the	66
the infringer was	66
as soon as	66
there is not	66
of information systems	66
of the value	66
tend to be	66
most of these	66
o n f	66
proposed method is	66
result in the	66
concerned with the	66
a l u	66
for the expert	66
to account for	66
order of the	66
limited number of	66
b la ck	66
e r m	66
and support vector	66
the feature selection	66
sent to the	65
e d m	65
examples of the	65
of the major	65
in an expert	65
was used as	65
expert system shells	65
the hidden layer	65
that is used	65
information provided by	65
of the association	65
of more than	65
mentioned in section	65
s a n	65
criteria decision making	65
number of training	65
the power of	65
d u j	65
u l a	65