A Study on Legibility with Pairwise Comparison
in Simultaneous Multilingual Display on Digital Signage
Takashi Kurushima
1
, Yoshiki Sakamoto
1
, Kimi Ueda
1
, Hirotake Ishii
1 a
, Hiroshi Shimoda
1 b
,
Rika Mochizuki
2
and Masahiro Watanabe
2
1
Graduate School of Energy Science, Kyoto University, Yoshida-hommachi, Sakyo-ku, Kyoto, Japan
2
Organization of Service Evolution Laboratories, NTT Group , Hikarinooka, Yokosuka-shi, Kanagawa-ken, Japan
Keywords: Digital Signage, Multi-linguistic, Pairwise Comparison.
Abstract:
Recently, digital signages are widely utilized to provide information for tourists in plural languages, but it takes
longer time to offer information in each language separately. Although simultaneous multilingual display is
expected to reduce the time for presentation, there is no guideline for simultaneous multilingual display to
keep its legibility. In order to estimate the effect of layout on legibility, the authors prepared contents which
have different layout factors and conducted evaluated the legibility of contents with Thurstone’s pairwise
comparison. Since the number of combinations in layout factors increases exponentially, the number of stimuli
was reduced using an orthogonal table. As a result, significant differences were found in some layout factors
and it was found that horizontal arrangement of images and sentences, left alignment in sentences, middle
margin from screen edge and medium-sized images made contents more legible.
1 INTRODUCTION
Foreign tourism has been getting popular all over
the world and the number of foreign sightseers has
been increasing. The signs in sightseeing places are
therefore often displayed in multilingual these days.
Thanks to recent advancement of information and
communication technology, the signs are digitalized
as digital signage and it is widely used because it
is easy to update displayed information and less ex-
pensive devices. Especially in Japan, the govern-
ment is promoting its utilization and function expan-
sion for the coming Tokyo Olympics and Paralympics
2020 (the Ministry of Internal Affairs and Communi-
cations, 2015). However, there has been no concrete
guidelines for multilingual display of digital signage.
At present, although some digital signages changes
the displayed information depending on the language,
it needs much time to display the information. Other
signages show information in multilingual at the same
time. It needs less time to show the information, how-
ever, it may cause difficulty to read because of lots of
texts are displayed in the screen. It is therefore neces-
a
https://orcid.org/0000-0002-5638-4862
b
https://orcid.org/0000-0002-5231-4955
sary to develop the guideline for legible multilingual
display.
The aim of this study is, therefore, to investigate
the layout factors which affects legibility of simul-
taneous multilingual display in order to establish the
guideline for multilingual display. In this study, var-
ious simultaneous multilingual formats were created
with combinations of various layout factors and they
were evaluated by a subject experiment. Since there
are actually huge number of combinations of layout
factors, the number of the display to be evaluated was
reduced with experimental design using orthogonal
table. In addition, the evaluation was done by Thur-
stone’s pairwise comparison to reduce the burden of
the evaluators.
There have been lots of studies which deal with
the relationship between display layout and legibil-
ity in various research fields. Dyson evaluated the
relationship between legibility and text layout on
screen and it was found that the number of characters
per line affected reading speed (Dyson, 2004). Do-
bres et al. investigated the legibility in reading at a
glance and the result showed that crowded texts re-
quired more processing time than isolated texts (Do-
bres et al., 2018). Grozdanovic et al. assessed the
legibility of video display unit depending on charac-
112
Kurushima, T., Sakamoto, Y., Ueda, K., Ishii, H., Shimoda, H., Mochizuki, R. and Watanabe, M.
A Study on Legibility with Pairwise Comparison in Simultaneous Multilingual Display on Digital Signage.
DOI: 10.5220/0008162101120119
In Proceedings of the 3rd International Conference on Computer-Human Interaction Research and Applications (CHIRA 2019), pages 112-119
ISBN: 978-989-758-376-6
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
ter/background color combination and demonstrated
that negative contrast was more legible than positive
contrast (Grozdanovic et al., 2016). Fu et al. proposed
six measurements of legibility on web site using con-
tent size, font size, icon, color contrast, density and
alignment points and found that these factors gave a
significant impact on web page legibility (Fu and Su,
2009). Xie et al. focused on the location of digital sig-
nage and found the relationship between maximum
viewing distance and observation angle (Xie et al.,
2007).
In multilingual contents, on the other hand, there
are studies on readability of multilingual websites and
they indicated that not only translation but also re-
designing were required to improve readability (Mi-
raz et al., 2016) (Hussain et al., 2017). Ogi et al. de-
veloped multilingual digital signage which automati-
cally changed its language by the user’s smart phone
using iBeacon communication and demonstrated that
most users felt the effectiveness (Ogi et al., 2016). Al-
humoud et al. observed banking interfaces in Arabic
and English and showed that the social and cultural
aspect influenced the interfaces (Alhumoud et al.,
2015). Although there have been some studies on
legibility and multilingual contents, there is no study
which focuses on the legibility of simultaneous mul-
tilingual display for digital signage.
2 EVALUATION METHOD OF
LEGIBILITY
As mentioned above, there are various factors which
affect the legibility of simultaneous multilingual dis-
play and the number of the combinations would be
huge. And it is actually impossible to evaluate the
legibility of so many combinations one by one. In
this chapter, the evaluation method using Thurstone’s
pairwise comparison and a design method for multi-
factorial experiment was introduced.
2.1 Thurstone’s Pairwise Comparison
Thurstone’s pairwise comparison is the method that
evaluators select one of two stimuli which are one of
the combination pairs of several objects to be evalu-
ated, and they repeat the selection for all the combi-
nations. By analysing the evaluators’ response, the
scale values of the objects can be quantitatively cal-
culated. As calculation is made from the percentage
of selection in each combination instead of evalua-
tors’ response, individual inconsistency (e.g. three-
way standoff in selection) does not matter in calcula-
tion. This method is simple because they watch only
a pair of stimuli at a time and intuitively answer the
better one. Thurstone’s pairwise comparison, there-
fore, has an advantage of less evaluators’ workload
comparing the method where they directly evaluate
several objects at the same time.
2.2 Experimental Design Method
In the case that objects to be evaluated include several
factors, number of the combinations of all the factors
increases exponentially and it is difficult to evaluate
all the combinations. Assuming that the interrelation
of the factors is small, the number of the combina-
tion can be reduced using orthogonal table. Table 1
is an example of L
16
(2
15
) orthogonal table. In arbi-
trary pair of factor of the orthogonal table, all com-
binations appear the same number of times. In case
of 15 factors with 2 levels each, the number of com-
binations is originally 32768 (2
15
), however, it can be
reduced to only 16 combinations by using the table. A
column of the table expresses two-level factor while
four-level factor can be expressed with two columns.
By using orthogonal table, although stimuli are multi-
factoral, the effect of a certain factor can be analyzed
with one-way ANOVA as the effects of other factors
are canceled.
By employing Thurstone’s pairwise comparison
and the experimental design method, the evaluation
of multilingual display layout with multiple factors
which has originally huge number of combinations
can be done.
3 OBJECTIVE AND METHOD OF
EVALUATION EXPERIMENT
3.1 Objective
The objective of the experiment is to confirm the ef-
fectiveness of the evaluation method mentioned in
Chapter 2 and to investigate the layout factors which
affect the legibility of simultaneous multilingual dis-
play. In the experiment, the participants repeated to
watch a pair of the contents with different layouts and
select more legible one. The scale value of each con-
tent was calculated from the answer of the selections
and the layout factor which affected the legibility was
deduced.
3.2 Layout Factors
There are various layout factors for display of dig-
ital signage such as font size and margin. In this
A Study on Legibility with Pairwise Comparison in Simultaneous Multilingual Display on Digital Signage
113
Table 1: L
16
(2
15
) orthogonal table.
Column
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
3 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1
4 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0
5 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1
6 0 1 1 0 0 1 1 1 1 0 0 1 1 0 0
7 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0
8 0 1 1 1 1 0 0 1 1 0 0 0 0 1 1
9 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
10 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0
11 1 0 1 1 0 1 0 0 1 0 1 1 0 1 0
12 1 0 1 1 0 1 0 1 0 1 0 0 1 0 1
13 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0
14 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1
15 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1
16 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0
study, however, the information to be presented was
fixed to three images and short explanation text in
four languages, Japanese, English, Korean and Chi-
nese (Traditional Chinese). The order of texts was
also fixed, Japanese, English, Korean and Chinese be-
cause Japanese and Chinese have similar characters
and they should be written in isolation. And the lay-
out factors to be considered were also limited to (i)
arrangement, (ii) margin, (iii) image block size, (iv)
spacing between languages and (v) alignment, as the
first trial for the evaluation experiment. Other forms
of information and other factors will be considered in
further study. The layout factors are explained in fol-
lowing paragraphs.
3.2.1 Arrangement
The arrangement is the factor which expresses the ar-
rangement of images and texts. The arrangement is
chosen as a layout factor because it decides the over-
all layout of content. It has two levels, vertical ar-
rangement and horizontal arrangement. The vertical
arrangement means that images are located at the top
and texts are below the images as shown in Figure 1.
while the horizontal arrangement means that the fig-
ures are located on the left and texts are on the right
as shown in Figure 2.
3.2.2 Margin
The margin is the factor which expresses the margin
between the images and texts, and the margin from
the edge of the screen. The margin is chosen as a lay-
out factor because it gives a crowded or uncrowded
impression to evaluators. It has four levels, 0%, 2.5%,
5% and 7.5%. For example, 2.5% means that the mar-
gin from the right and left edges is 2.5% of the width
Figure 1: Example of Vertical Arrangement.
Figure 2: Example of Horizontal Arrangement.
of the screen and that that from the top and bottom is
also 2.5% of the height. The margin between the im-
ages and texts is set as well. The examples of layouts
with 0% and 7.5% margin are shown in Figure 3 and
Figure 4.
3.2.3 Image Block Size
The image block size is the factor which indicates the
ratio of image area among the whole screen area. The
image block size is chosen as a layout factor to assess
the impact of images. It has four levels, 20%, 30%,
40% and 50%. When the image block size is 30%, for
example, the height of the images is set to 30% of the
CHIRA 2019 - 3rd International Conference on Computer-Human Interaction Research and Applications
114
Figure 3: Example of 0% Margin.
Figure 4: Example of 7.5% Margin.
height of the screen when angement while the width
of the images is set to 30% of the screen when hor-
izontal arrangement. The examples of layouts with
20% and 50% image block size are shown in Figure 5
and Figure 6. In case that the image overlaps the mar-
gin mentioned in 3.2.2, it is trimmed or downsized.
Figure 5: Example of 20% Image Block Size.
Figure 6: Example of 50% Image Block Size.
3.2.4 Spacing between Languages
The spacing between languages is the factor which
indicates the margin between the explanation texts in
four languages. The spacing between languages is
chosen as a layout factor because finding evaluator’s
language may be important in simultaneous multilin-
gual display. It has four levels, 0%, 50%, 100% and
150%. For example, 50% means that the margin be-
tween the texts in Japanese and those in English is set
to 50% of font height. The examples of layouts with
0% and 150% spacing between languages are shown
in Figure 7 and Figure 8. The font size is automati-
cally set to the maximum size under the limitation of
margin, image block size and spacing between lan-
guages.
Figure 7: Example of 0% Spacing between Languages.
Figure 8: Example of 150% Spacing between Languages.
3.2.5 Alignment
The alignment is the factor which shows the align-
ment of the texts. The alignment is chosen as a lay-
out factor because it may affect legibility of text. It
has two levels, left alignment and center alignment.
The examples of layouts with left alignment and cen-
ter alignment are shown in Figure 9 and Figure 10.
3.3 Presented Information
Three kinds of information were prepared for the ex-
periment, which were (a) sightseeing information, (b)
A Study on Legibility with Pairwise Comparison in Simultaneous Multilingual Display on Digital Signage
115
Figure 9: Example of Left Alignment.
Figure 10: Example of Center Alignment.
disaster evacuation information and (c) local food cul-
ture . They were chosen assuming the case that the
digital signage was placed in the sightseeing places
for foreign tourists.
In each information, 16 sorts of layout were pre-
pared along with the L
16
orthogonal table and 5 layout
factors above. The layout factors and prepared con-
tents are shown in Table 2. The examples of contents
are shown in Figure 11-13.
Figure 11: Example of Sightseeing Information (#13); Pic-
tures are (Co0316co, 2016a) (Co0316co, 2016b) (KENPEI,
2008).
3.4 Experimental Procedure
Table 3 shows the experimental procedure. In each
evaluation of “Evaluation 1-3”, a pair of layouts were
displayed on two monitors (42.5 inches LCD, LG
43UD79-B, resolution: 3840x2160) right and left, re-
spectively, and the participants selected more legible
one by using keyboard. Since the number of com-
Figure 12: Example of Disaster Evacuation (#8).
Figure 13: Example of Local Food Culture (#11).
binations of 2 layouts among 16 layouts is 120, the
selections were repeated 120 times for each presented
information. The display time for the selection was
set to 6.5 seconds, while the time between the selec-
tions was 2.0 seconds. In this experiment, as partic-
ipants evaluated the legibility at a glance, layout is
more effective in legibility than the detail of contents
or participants’ background. The distance between
the monitor and the participants was 2.0m. In “Eval-
uation 1-3” in Table 3, the order of three informa-
tion, (a) sightseeing information, (b) disaster evacu-
ation information and (c) local food culture were ran-
domly assigned to each participant in order to cancel
the order effect. In “Practice” in Table 3, the partici-
pants practiced the selection using different informa-
tion and layouts from those of “Evaluation 1-3.
3.5 Participants
25 participants joined the experiment who had nor-
mal vision with or without correction and their native
language was Japanese. They were 15 males and 10
females and their average age was 21.6 (SD: 1.69).
CHIRA 2019 - 3rd International Conference on Computer-Human Interaction Research and Applications
116
Table 2: The layout of contents.
Factor (The numbers in round bracket mean corresponding columns in Table 1)
Arrangement Margin Image Block Size Spacing between languages Alignment
# (6) (1, 12) (4, 11) (3, 8) (5)
1 Vertical 0% 20% 0% Center
2 Vertical 2.5% 30% 50% Center
3 Horizontal 2.5% 50% 0% Left
4 Horizontal 0% 40% 50% Left
5 Horizontal 0% 30% 150% Center
6 Horizontal 2.5% 20% 100% Center
7 Vertical 2.5% 40% 150% Left
8 Vertical 0% 50% 100% Left
9 Vertical 7.5% 30% 150% Left
10 Vertical 5% 20% 100% Left
11 Horizontal 5% 40% 150% Center
12 Horizontal 7.5% 50% 100% Center
13 Horizontal 7.5% 20% 0% Left
14 Horizontal 5% 30% 50% Left
15 Vertical 5% 50% 0% Center
16 Vertical 7.5% 40% 50% Center
Table 3: Procedure of experiment.
Time (Min.) Procedure
10 Introduction and Questionnaire
5 Practice
20 Evaluation 1
5 Break
20 Evaluation 2
5 Break
20 Evaluation 3
10 Final procedure
4 RESULTS OF EVALUATION
EXPERIMENT
In order to assess the impact of layout factors, scale
value of each content was calculated from partici-
pants’ response and the effect of each factor was an-
alyzed with one-way ANOVA. In this experiments,
there are two analysis methods which are (1) the case
that the answers of all the participants are analyzed at
a time and (2) the case that the order effect of three
sorts of information for “Evaluation 1-3” is consid-
ered. The analysis results of (1) and (2) are described
below;
4.1 Analysis Result for All Participants
In this section, the analysis results of each presented
information for all the participants are described.
First, the selection of all the participants were accu-
mulated and the scale values of 16 layouts were cal-
culated by Thurstone’s pairwise comparison method.
Then, the scale values of each level of the layout fac-
tors were compared by statistical method and p val-
ues were calculated to figure out the significant dif-
ference.
4.1.1 Sightseeing Information
For the arrangement, the horizontal arrangement was
evaluated more legible than the vertical arrangement,
though there is no significant difference (F(1, 14) =
1.42, p = 0.25). For the margin, the legible or-
der was 5%, 2.5%, 7.5% and 0%, though there is
no significant difference either (F(3,12) = 0.93, p =
0.46). For the image block size, the legible order
was 30%, 40%, 50% and 20% and significant dif-
ference was found (F(3, 12) = 5.92, p = 0.01). A
post hoc Tukey test showed that 30% and 40% were
significantly more legible than 20% (p < 0.05). For
the spacing between languages, the legible order was
100%, 150%, 0% and 50%, though there was no sig-
nificant difference (F(3, 12) = 0.15, p = 0.93). For
the alignment, left alignment was evaluated to be
more legible, though there is no significant difference
either (F(1, 14) = 1.04, p = 0.33).
4.1.2 Disaster Evacuation Information
For the arrangement, the horizontal arrangement was
evaluated more legible than the vertical arrange-
ment and significant tendency was found (F(1, 14) =
3.16, p = 0.097). For the margin, the legible order
was 5%, 2.5%, 0% and 7.5%, though there was no
significant difference (F(3, 12) = 2.56, p = 0.10). For
the image block size, the legible order was 30%, 40%,
50% and 20%, though there was no significant differ-
ence either (F(3, 12) = 1.49, p = 0.27). For the spac-
ing between languages, the legible order was 50%,
A Study on Legibility with Pairwise Comparison in Simultaneous Multilingual Display on Digital Signage
117
100%, 150% and 0%, though there is no significant
difference either (F(3, 12) = 0.03, p = 0.99). For the
alignment, left alignment was evaluated to be more
legible, though there is no significant difference either
(F(1, 14) = 0.12, p = 0.73).
4.1.3 Local Food Culture
For the arrangement, the horizontal arrangement was
evaluated more legible than the vertical arrange-
ment and significant tendency was found (F(1, 14) =
3.20, p = 0.09). For the margin, the legible order was
5%, 2.5%, 0% and 7.5%, though there is no signifi-
cant difference (F(3, 12) = 0.75, p = 0.54). For the
image block size, the legible order was 30%, 40%,
50% and 20% and significant difference was found
(F(3, 12) = 3.77, p = 0.04), though a post hoc Tukey
test didn’t show any significant difference between
levels. For the spacing between languages, the legible
order was 100%, 150%, 50% and 0%, though there is
no significant difference (F(3, 12) = 0.08, p = 0.97).
For the alignment, left alignment was evaluated to be
more legible, though there is no significant difference
either (F(1, 14) = 1.77, p = 0.20).
4.2 Analysis Result Considering Order
Effect
Three presented information such as (a) sightsee-
ing information, (b) disaster evacuation information
and (c) local food culture were randomly assigned
to “Evaluation 1-3” in the experiment and the order
effect of the evaluation might affect the results. In
this section, therefore, the analysis results consider-
ing the order effect of the presented information are
described. Actually the participants were divided into
3 groups depending on the order and the scale values
of each layout were calculated separately. Then, sta-
tistical comparison was done using all the scale val-
ues.
4.2.1 Sightseeing Information
For the arrangement, the horizontal arrangement was
evaluated more legible than the vertical arrange-
ment and significant tendency was found (F(1, 46) =
3.84, p = 0.06). For the margin, the legible order was
5%, 2.5%, 7.5% and 0% and significant tendency was
found too(F(3, 44) = 2.72, p = 0.06), though a post
hoc Tukey test didn’t show any significant difference
between levels. For the image block size, the legible
order was 30%, 40%, 50% and 20% and significant
difference was found (F(3, 44) = 18.35, p < 0.01). A
post hoc Tukey test showed that 30% and 40% were
significantly more legible than 50% and 20% (p <
0.05). For the spacing between languages, the legible
order was 100%, 150%, 0% and 50%, though there
was no significant difference (F(3, 44) = 0.28, p =
0.84). For the alignment, left alignment was evalu-
ated to be more legible and significant tendency was
found (F(1, 46) = 2.88, p = 0.097).
4.2.2 Disaster Evacuation Information
For the arrangement, the horizontal arrangement was
evaluated more legible than the vertical arrangement
and significant difference was found (F(1, 46) =
10.39, p < 0.01). For the margin, the legible order
was 5%, 2.5%, 0% and 7.5%, and significant differ-
ence was found too (F(3, 44) = 6.45, p < 0.01). A
post hoc Tukey test showed that 5% and 2.5% were
significantly more legible than 7.5% (p < 0.05). For
the image block size, the legible order was 30%, 40%,
50% and 20%, and significant difference was found
too (F(3, 44) = 5.49, p < 0.01). A post hoc Tukey
test showed that 30% was significantly more legi-
ble than 50% and 20% (p < 0.05). For the spac-
ing between languages, the legible order was 100%,
50%, 150% and 0%, though there is no significant
difference (F(3, 44) = 0.17, p = 0.92). For the align-
ment, left alignment was evaluated to be more legible,
though there was no significant difference (F(1, 46) =
0.61, p = 0.44).
4.2.3 Local Food Culture
For the arrangement, the horizontal arrangement was
evaluated more legible than the vertical arrangement
and significant difference was found (F(1, 46) =
7.86, p < 0.01). For the margin, the legible order was
5%, 2.5%, 0% and 7.5%, though there is no signifi-
cant difference (F(3, 44) = 2.10, p = 0.11). For the
image block size, the legible order was 30%, 40%,
50% and 20% and significant difference was found
(F(3, 44) = 10.81, p < 0.01). A post hoc Tukey test
showed that 30% and 40% was significantly more leg-
ible than 50% and 20% (p < 0.05). For the spacing
between languages, the legible order was 100%, 50%,
150% and 0%, though there is no significant differ-
ence (F(3, 44) = 0.17, p = 0.92). For the alignment,
left alignment was evaluated to be more legible, and
significant difference was found (F(1, 46) = 4.27, p =
0.044).
5 DISCUSSION
The tendencies of each layout factor and their reasons
are discussed in this chapter. As the evaluation re-
CHIRA 2019 - 3rd International Conference on Computer-Human Interaction Research and Applications
118
sult of the arrangement, the horizontal arrangement
was more legible and significant difference was found
in some analysis results. The reason was supposed
that the length of lines in horizontal arrangement was
shorter than that in vertical arrangement and it looked
easy to read. As the result of the margin, 2.5% and
5% tended to be more legible and significant tendency
appeared in some results. It was supposed that appro-
priate margin was necessary to make the layout legi-
ble. As the result of image block size, 30% and 40%
were evaluated more legible and significant difference
was often found. In case of multilingual contents, the
amount of the text becomes larger. It was therefore
supposed that the ratio of image area should be a little
smaller than that of text for legible layout. As the re-
sult of alignment, the left alignment was more legible
and significant difference was found in some results.
It was supposed that the start position of reading was
always left edge in the left alignment and it gave leg-
ible impression.
6 CONCLUSION
Aiming to investigate the layout factors for legible
simultaneous multilingual display in this study, the
authors have proposed the evaluation method which
employs Thurstone’s pairwise comparison and exper-
imental design using orthogonal table, and a trial ex-
periment was conducted using only the limited dis-
play content with three images and explanation texts
in four languages. As the results, horizontal arrange-
ment with images and text, appropriate margin from
the edge of screen and left alignment of the text were
evaluated to be legible layout factors.
In this experiment, there are some limitations.
Since the display contents were limited to three im-
ages and explanation texts, the results cannot be ex-
panded to general layout. It is therefore necessary to
conduct other evaluation for general display contents.
And the native language of all the participants was
Japanese in this experiment. It is necessary to conduct
the same experiment for other participant with dif-
ferent native language to find the difference depend-
ing on the languages. In addition, this study did not
take readability or perspicuity into account because
the participants evaluated the legibility in a short time.
Future works can consider reading speed and compre-
hension in simultaneous multilingual display.
When revealing the layout factors of legible si-
multaneous multilingual display in the future, it is ex-
pected to realize an automatic layout tools for simul-
taneous multilingual display can be developed. Using
such tools, even the person who has no special knowl-
edge and skill can easily create legible simultaneous
multilingual display contents.
REFERENCES
Alhumoud, S., Alabdulkarim, L., Almobarak, N., and Al-
Wabil, A. (2015). Socio-cultural aspects in the design
of multilingual banking interfaces in the arab region.
In Kurosu, M., editor, Human-Computer Interaction:
Users and Contexts, pages 269–280, Cham. Springer
International Publishing.
Co0316co (2016a). The fushimi inari there are many
torii. CC-BY-SA 4.0 (https://creativecommons.org/
licenses/by-sa/4.0/).
Co0316co (2016b). The image of the fox that ex-
ample mouth the jewel. CC-BY-SA 4.0 (https://
creativecommons.org/licenses/by-sa/4.0/).
Dobres, J., Wolfe, B., Chahine, N., and Reimer, B. (2018).
The effects of visual crowding, text size, and posi-
tional uncertainty on text legibility at a glance. Ap-
plied Ergonomics, 70:240 – 246.
Dyson, M. C. (2004). How physical text layout affects read-
ing from screen. Behaviour & Information Technol-
ogy, 23(6):377–393.
Fu, F.-L. and Su, C.-H. (2009). Formalizing design guide-
lines of legibility on web pages. In Salvendy, G. and
Smith, M. J., editors, Human Interface and the Man-
agement of Information. Information and Interaction,
pages 17–25, Berlin, Heidelberg. Springer Berlin Hei-
delberg.
Grozdanovic, M., Marjanovic, D., Janackovic, G., and
Djordjevic, M. (2016). The impact of charac-
ter/background colour combinations and exposition
on character legibility and readability on video display
units. Transactions of the Institute of Measurement
and Control, 39.
Hussain, W., Hussain, O. K., Hussain, F. K., and Khan,
M. Q. (2017). Usability evaluation of english, local
and plain languages to enhance on-screen text read-
ability: A use case of pakistan. Global Journal of
Flexible Systems Management, 18(1):33–49.
KENPEI (2008). Fushimi Inari-taisha Gehaiden. CC-
BY-SA 3.0 (https://creativecommons.org/licenses/by-
sa/3.0/).
Miraz, M. H., Excell, P. S., and Ali, M. (2016). User inter-
face (ui) design issues for multilingual users: a case
study. Universal Access in the Information Society,
15(3):431–444.
Ogi, T., Ito, K., and Konita, S. (2016). Multilingual digital
signage using ibeacon communication. In 2016 19th
International Conference on Network-Based Informa-
tion Systems (NBiS), pages 387–392.
the Ministry of Internal Affairs and Communications
(2015). Action plan on ict of the whole society to-
wards 2020 (1st edition).
Xie, H., Filippidis, L., Galea, E. R., Gwynne, S., Black-
Shields, D., and Lawrence, P. J. (2007). Experimental
study and theoretical analysis of signage legibility dis-
tances as a function of observation angle. In Waldau,
N., Gattermann, P., Knoflacher, H., and Schrecken-
berg, M., editors, Pedestrian and Evacuation Dynam-
ics 2005, pages 131–143, Berlin, Heidelberg. Springer
Berlin Heidelberg.
A Study on Legibility with Pairwise Comparison in Simultaneous Multilingual Display on Digital Signage
119