INFLUENCING FACTORS OF HIGH-SPEED RAILWAY
PASSENGERS’ TRAVEL CHOICE BASED ON ROUGH SET
Fan Yuhang
1
, Li Jing
2
1
School of Economics and Management, Beijing Jiaotong University
Shangyuan Road 3, Haidian District, 100044, Beijing, China
2
School of Economics and Management, Beijing Jiaotong University
Shangyuan Road 3, Haidian District, 100044, Beijing, China
Long Chenxu
3
3
School of Economics and Management, Beijing Jiaotong University
Shangyuan Road 3, Haidian District, 100044, Beijing, China
Keywords: High-speed railway, Passenger travel, Rough set, Combination weight.
Abstract: Inter-city traffic is a large complicated system, solving the inter-city traffic problems lies in the reasonable
allocation of the methods of transportation, and its core is to satisfy the need of high-speed railway passenger
travel. By using rough set theory, this paper calculated the single factor weight and multi-motion weights for
the 5 main influencing factors of High-speed railway passengers’ choice of travel, and cleared out the
interaction and contacts between the various influencing factors, provided scientific decision support for the
construction of high-speed railway and the actual operation management.
1 INTRODUCTION
Nowadays, aviation, high-speed railway, inter-city
railway, highway have a fierce market competition.
In this situation, passengers’ choice of travel will be
related to operation efficiency of the different ways
of transportation directly. With the development of
social economy and people travel more and more
frequently, how to satisfy people's demand for travel
in the best way of transportation and the most
reasonable arrangement is the most important
problems of high-speed railway operation, also one
important guiding principle of high-speed railway’s
development. But the existing research of
passengers’ travel choice are most qualitative
analysis of behavioral characteristics at specific
travel stage, Quantitative research is limited to that
such as PDL passenger share rate estimation, The
share rate of research are most based on the random
utility theory analysis, the division of the way of
transportation is a basic work, and no one has
determine the importance degree of the influencing
factors of travel choice. Therefore, this paper applied
the rough set theory method, established the model
to calculate the single factor weight and multi-
motion weights of high-speed railway passengers’
travel choice, and revealed the mechanism of traffic
modes’ selection when high-speed railway passenger
travel, provide scientific decision support for the
construction of high-speed railway and the actual
operation management.
2 COMBINATION WEIGHT OF
MULTIPLE ALGORITHM
BASED ON ROUGH SET
According to rough set principle, a knowledge
representation system S could be built by describing
research objects’ attributes and attribute values, that
is
(
)
,,,)SUCDVf=∪
(1)
213
Yuhang F., Jing L. and Chenxu L..
INFLUENCING FACTORS OF HIGH-SPEED RAILWAY PASSENGERS’ TRAVEL CHOICE BASED ON ROUGH SET.
DOI: 10.5220/0003552002130217
In Proceedings of the 13th International Conference on Enterprise Information Systems (ICEIS-2011), pages 213-217
ISBN: 978-989-8425-54-6
Copyright
c
2011 SCITEPRESS (Science and Technology Publications, Lda.)
In this equation,
{
}
12
,,Uxx=
represents the set of
research objects;
12
{, , }Ccc=
and
12
{, , }Ddd=
represent the sets of research objects’ condition
attributes and decision-making attributes, in addition
we have
ADC =
and
CD
φ
∩=
;V is the set of
attribute values;
:
f
UA V×→
is an information
function, it endows every research object’s each
attribute with an
information value.
The importance of a condition attribute
i
c
could be
expressed as
[()]
() 1
[()]
i
i
Cc
Cc i
C
Card Pos D
Sig c
Card Pos D
=−
(2)
In this equation,
i
c
C; i=1,2,3,…,n; n
represents the number of elements in the condition
attribute set;
()
C
P
os D
represents the object set in
which all elements from U could be divided into the
equivalent categories of relation D according to the
classification information
/( )
i
UCc
;
[ ]Card
represents the set’s cardinal number, which means
the number of elements in the set.
The weight of every condition attribute could be
gained through executing normalized process on the
importance of each according to equation (2), which
is
1
()
()
i
i
Cc i
c
n
Ck
k
Sig c
W
Sig c
=
=
(3)
According to the definition of Rough Set principle,
equation (2) and (3) could be extended to get the
combination weight of several condition
attributes
,,
ij
cc
, namely the combination weight
of multiple, which represents the influences on
decision-making result when multiple condition
attributes combine and interact on each other.
(,, )
(,, )
()
1
(, , )
(,1,2,,,,,)
()
ij
ij
ij
Cc c i j
cc
nn
Cc c i j
iji
Sig c c
Wijinjin
Sig c c
==
=≥==
∑∑

(4)
Besides,
(,, )
(,, )
[()]
(, , ) 1
[()]
ij
ij
Cc c
Cc c i j
C
Card Pos D
Sig c c
Card Pos D
=−
(5)
In this equation,
(,, )
(, , )
ij
Cc c i j
Sig c c
represents the
magnitude of the combination of several condition
attributes;
(,, )
(, , )
ij
Cc c i j
Pos c c
represents the set of
all the objects which could be accurately divided
into equivalent categories of relation D according to
the classification information
/(,,)
ij
UC c c
.
According to equation (4) and (5), two or more
condition attributes could be combined together,
thus the combination weight of multiple can be
calculated applying basic theory of Rough Set
principle.
The biggest advantage of adopting the calculation
method of combination weight of multiple based on
rough set to analyze the combination of several
condition attributes’ influences on systems’
decision-making is that not only single attribute’s
effects have been taken into consideration, but the
interactions among attributes and their influences on
decision-making have also been considered, so one-
sidedness and limitations of calculating one single
attribute’s weight could be avoided. Therefore, this
method is especially fit for the research on
influencing factors of High-speed railway
passengers’ travel choice.
3 THE EXAMPLE OF
INFLUENCING FACTORS OF
HIGH-SPEED RAILWAY
PASSENGERS’ TRAVEL
CHOICE BASED ON ROUGH
SET
3.1 Factors’ Selecting
High-speed railway system consists of High-speed
railway operation, High-speed railway facilities and
High-speed railway management policy. As a
subsystem of the transportation system, it’s most
important function is to satisfy passengers’ travel
demand. Therefore, this paper used the concept that
high-speed railway is service in "person" as a
starting point, and the influencing factors of
passengers’ travel choice as the breakthrough point,
to research high-speed railway and the influencing
factors of passengers’ travel choice. Through data
analysis and investigation, this paper determined
high-speed railway passengers’ travel choice index
system, including five influencing factors of high-
speed railway passengers’ travel choice and one
factors for evaluation of the result, as table 1 shows.
Among them, the five influencing factors are the
ticket price satisfaction, the train’s start time
satisfaction, the train’s speed satisfaction, the taking
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
214
environment satisfaction, the high-speed railway
safety level, one factors for evaluation of the result
is the overall satisfaction for high-speed railway.
3.2 Knowledge Expression of
Influencing Factors
As table 1 shows, influencing factors of High-speed
railway passengers’ travel choice not only include
the Humanness co-constructs factors such as time
and price, but also the comfort level and safety.
Based on rough sets theory, taking the 863 passenger
questionnaire record at the Spring Festival period in
2009 as a research collections of objects, setting
as U={x
1
,x
2
,x
3
,x
4
,x
5
...x
861
,x
862
,x
863
}, then using
c
1
c
2
c
3
c
4
c
5
to represent the ticket price
satisfaction, the train’s start time satisfaction, the
train’s speed satisfaction, the taking environment
satisfaction and the high-speed railway safety level,
And constituting condition attributes set of
influencing factors of High-speed railway
passengers’ travel choice as C={ c
1
c
2
c
3
c
4
c
5
},
meanwhile, using the overall satisfaction d to
constitute decision attribute set D. Thus,
constructing a two-dimensional decision-making
information table to describe influencing factors of
High-speed railway passengers’ travel choice
knowledge expression system, table 2 shows part of
the constructed decision table.
Table 1: High-speed railway passengers’ travel choice index system.
Target Criterion Index
Influencing Factors of
High-speed Railway Passengers’
Travel Choice
Passengers Characteristics
Sex
Age
Education Degree
Vocation
Monthly Income
Price
Travel Expense Bearing Way
The Ticket Price Satisfaction
If fares improve will choose of transportation
Ticket Price
Taking Time
The train’s start time satisfaction
Running time increases will choose of transportation
The Train’s Speed Satisfaction
Time to the Station
Time for Ticket
Time for Waiting the Train
Time for Travelling
Distance
Time for Midway Trasfer
Time from Station to House
Taking environment
The Taking Environment Satisfaction
The High-speed Railway Safety Level
The Overall Satisfaction
Taking Type
INFLUENCING FACTORS OF HIGH-SPEED RAILWAY PASSENGERS' TRAVEL CHOICE BASED ON ROUGH
SET
215
Table 2: Influencing factors of High-speed railway passengers’ travel choice knowledge expression system.
number
Condition attribute
Decision attribute
c1. the ticket
price
satisfaction
c2. the train’s
start time
satisfaction
c3. the train’s
speed
satisfaction
c4. the taking
environment
satisfaction
c5. the high-
speed railway
safety level
d. the overall
satisfaction
1 3 5 2 4 5 5
2 5 4 4 4 4 4
3 3 3 2 5 5 5
4 4 4 4 4 5 5
5 4 4 4 4 4 4
6 2 4 1 3 6 4
7 3 4 3 5 6 3
8 3 4 1 4 4 4
9 2 4 6 5 6 6
10 4 3 1 5 2 4
... ... ... ... ... ... ...
862 3 3 2 6 5 4
863 2 4 2 3 5 3
3.3 CALCULATION
3.3.1 The Actual Calculation Results of
Single Influencing Factors of
High-speed Railway Passengers’
Travel Choice Shows in Table 3
Table 3: Single influencing factors of High-speed railway
passengers’ travel choice.
Influencing Factor Importance Degree Weight
c1. the ticket price
satisfaction
0.251 0.196
c2. the train’s start time
satisfaction
0.267 0.207
c3. the train’s speed
satisfaction
0.273 0.213
c4. the taking
environment satisfaction
0.263 0.205
c5. the high-speed
railway safety level
0.231 0.179
3.3.2 Multi-motion Weights Shows in
Table 4
Table 4: Multi-motion weights.
Influencing factor
Importance
degree
Weight
c
1
, c
2.
the ticket price satisfaction
the
train’s start time satisfaction
0.554 0.0996
c
1
, c
3.
the ticket price satisfaction
the
train’s speed satisfaction
0.552 0.0991
c
1
, c
4.
the ticket price satisfaction
the
taking environment satisfaction
0.533 0.0958
c
1
, c
5.
the ticket price satisfaction
the
high-speed railway safety level
0.519 0.0933
c
2
, c
3.
the train’s start time satisfaction
the
train’s speed satisfaction
0.586 0.1054
c
2
, c
4.
the train’s start time satisfaction
the
taking environment satisfaction
0.585 0.1052
c
2
, c
5.
the train’s start time satisfaction
the
high-speed railway safety level
0.550 0.0989
c
3
, c
4.
the train’s speed satisfaction
the
taking environment satisfaction
0.587 0.1056
c
3
, c
5.
the train’s speed satisfaction
the
high-speed railway safety level
0.555 0.0998
c
4
, c
5.
the taking environment
satisfaction
the high-speed railway safety
level
0.541 0.0973
ICEIS 2011 - 13th International Conference on Enterprise Information Systems
216
3.3.3 Calculation Result Analysis
The actual calculation results of single influencing
factors of High-speed railway passengers’ travel
choice shows, by large to small, the influencing
factors weights of travel choice are the train’s speed
satisfaction, the train’s start time satisfaction, the
taking environment satisfaction, the ticket price
satisfaction and the high-speed railway safety level.
This paper uses the original data from the Spring
Festival transportation period questionnaire.
Passengers in this period are most students to return
hometown and working-class that work outside. For
the eager to go home for the Spring Festival, they
have high requirements on the travel speed that the
time they have to take to get home; they also have
high requirements on the train’s start time that
whether the train is late. As the saying goes that ” be
thrifty at home and spend liberally while travelling”.
They also hope to get a better taking comfort.
Relatively speaking, in the Spring Festival period,
passengers are not too sensitive to the ticket price
and safety.
The actual calculation results of Multi-motion
weights shows, the Combination weights of speed
and taking environment is the highest, then speed
and start time. the Combination weights of start time
and taking environment is third. The result indicates
that in the Spring Festival period, passengers are
satisfied to the shorter time and the better taking
environment mostly, the union is the main
determinants that passengers choose high-speed
railway.
4 CONCLUSIONS
According to Rough Set theory, we could not only
calculate one single attribute’s weight but the
combination weights of two or more of them. To
combine the calculations of single weight and
combination of multiple for comparison and
analysis, we could analyze the condition attributes
themselves and their relationships more accurately
and objectively. Thus, the Rough Set Theory is
extraordinarily appeal to the analysis of factors
affecting High-speed Railway passengers’ travel
choices.
The calculation results of one factor or the
combination of two or more of them that affects
High-speed Railway passengers’ choices of travel
shows that High-speed railway passengers regard the
speed as the most important factors when choosing
travel mode. Next to speed are the time of departure
and the environment. Though the weight of single
factor of departure time is higher than that of the
environment, the calculation of combination of
multiple factors shows that the combination of
railway speed and environment influence the most in
passengers’ travel choices.
REFERENCES
Fan Zhi-ping, Zhang Quan, Ma Jian.1998. An integrated
approach to determining weights in multiple attribute
decision making . Journal of Management Sciences in
China. 1998,1 ( 3) : 50-53.
Pawlak Z.1991. Rough Sets: Theoretical aspects of
reasoning about data . Boston, Kluwer Academic
Publishers. USA., 1
st
edtion.
Hao Cheng, Li Jing, Li Xue-mei, Li Xue-wei.2008..
Research on the combination weight of multiple
benefit factor s in urban rail transit project based on
rough set. China Railway Science. 2008,11(6):121-
125.
Miao Duo-qian, Wang Jue.1999. An information
representation of the concepts and operations in rough
set theory. Journal of Software. 1999,10 ( 2) : 113-116
Xue Zhan-jun.2008. The strategic significance of chinese
high-speed railway’s development. Science and
Technology Innovation Herald. 2008,(21):55
INFLUENCING FACTORS OF HIGH-SPEED RAILWAY PASSENGERS' TRAVEL CHOICE BASED ON ROUGH
SET
217