A Study on Legibility with Pairwise Comparison

in Simultaneous Multilingual Display on Digital Signage

Takashi Kurushima

, Yoshiki Sakamoto

, Kimi Ueda

, Hirotake Ishii

1 a

, Hiroshi Shimoda

1 b

Rika Mochizuki

and Masahiro Watanabe

Graduate School of Energy Science, Kyoto University, Yoshida-hommachi, Sakyo-ku, Kyoto, Japan

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-

https://orcid.org/0000-0002-5638-4862

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

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

) 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

), 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

) 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

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