PROPOSAL OF A FRAMEWORK TO SHARE KNOWLEDGE

ON CONSUMER’S IMPRESSIONS

Keiichi Muramatsu

∗

Graduate School of Human Sciences, Waseda University, Shinjuku, Tokyo, Japan

Tatsuo Togawa

Advanced Research Center for Human Sciences, Waseda University, Shinjuku, Tokyo, Japan

Kazuaki Kojima, Tatsunori Matsui

Faculty of Human Sciences, Waseda University, Shinjuku, Tokyo, Japan

Keywords:

Impression, Awareness, Aesthetic experience, Ontology engineering.

Abstract:

Recently, impressions of web pages formed by visitors have become an important tool to increase the number

of repeat visitors to the web page. Therefore, the management of knowledge on consumers’ impressions

obtained in several study ﬁelds is an essential task in current industrial design. However, no methods that

realize such knowledge management have been established. Thus, our study proposes and implements a

knowledge management method that can effectively provide knowledge of impressions to web designers to

help them in building attractive websites. We introduce a framework for the description of impressions that

depend on perceptual ﬂuency, which can serve as an useful indicator of pleasure. We can extract the features

of objects that affect impressions on the basis of perceptual ﬂuency. We specify the relationship between

objects and impressions by modeling the concepts of awareness, perception, and self-report on the basis of

an ontology development environment Hozo and a top-level ontology YAMATO. We then instantiate a case

where a person has a good impression of a Web page, and we describe the relationship between a perception

and a stimulus in such a case. Our approach demonstrates that ontological modeling of impressions helps us

to understand the correspondences between affections and physical irritations.

1 INTRODUCTION

1.1 Impressions of Web Pages

Consumers’ impression is one of the important fac-

tors in current industrial design. Manufacturers are

constantly looking for ideas to make a good impres-

sion on consumers, because products that are liked by

consumers may naturally result in an increase in sales.

This fact is true in the domain of Web design as well

in the sense that visitors to Web sites can be regarded

as consumers of Web pages. If a consumer forms a

good impression of the pages contained in a Web site,

he or she may visit the Web site again. A good ﬁrst

impression is the key to get more repeat visitors, and

∗

Research Fellow of the Japan Society for thePromotion

of Science (JSPS)

the ﬁrst impression based on the visual appeal of Web

pages can be formed within 50 ms (Lindgaard et al.,

2006). Therefore, for Web designers, management

of knowledge on how to inﬂuence consumers at the

ﬁrst glance is an important aspect of building attrac-

tive Web sites, which in turn can generate more trafﬁc.

In various research ﬁelds, researchers have studied

the relationship between characteristics of objects and

elements, such as color and shape, and the impression

formed by the viewer. While ﬁndings in these studies

are closely related to each other, these ﬁndings remain

isolated to their speciﬁc study ﬁelds. By systemati-

cally organizing these ﬁndings, useful knowledge that

can be shared across ﬁelds can be obtained. For ex-

ample, ﬁndings of impressions created by two color

combinations in psychology are useful for facade de-

to collectively manage such knowledge.

388

Muramatsu K., Togawa T., Kojima K. and Matsui T..

PROPOSAL OF A FRAMEWORK TO SHARE KNOWLEDGE ON CONSUMER’S IMPRESSIONS.

DOI: 10.5220/0003184203880393

In Proceedings of the 3rd International Conference on Agents and Artiﬁcial Intelligence (ICAART-2011), pages 388-393

ISBN: 978-989-8425-40-9

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

1.2 Perceptual Fluency

Lindgaard et al. (2006) studied visual appeal from

the viewpoint of aesthetics; studies on aesthetic expe-

rience can provide helpful insights into understand-

ing the impressions formed by consumers. Aesthetic

pleasure is explained by perceptual ﬂuency, which is

deﬁned as the ease with which stimuli can be phys-

ically identiﬁed (e.g., symmetry, contrast, prototypi-

cality); ﬂuent processing emotionally evokes positive

responses to the stimuli (Reber et al., 2004). For ex-

ample, a ﬁgure with high-contrast components is ex-

pected to evoke high ﬂuency, and the stimulus that

is ﬂuently perceived elicits a positive reaction from

the perceiver. Therefore, perceptual ﬂuency can serve

as a useful indicator for exploring impressions. We

can see the effects of perceptual ﬂuency in the early

stages of a model of aesthetic appreciation (Leder

et al., 2004). This model has ﬁve stages (perceptual

analyses, implicit memory integration, explicit clas-

siﬁcation, cognitive mastering, and evaluation), and

the ﬁrst two stages are related to perceptual ﬂuency.

In the stage of perceptual analyses, processing of per-

ceptual variables such as contrast, complexity, color,

and symmetry proceeds quickly, and in the stage of

implicit memory integration, aesthetic preferences are

affected by features such as familiarity and prototyp-

icality. In this paper, we regard the positive reactions

of perceivers facilitated by perceptual ﬂuency during

the early stages of human informational processing as

a measure of their impressions of Web pages.

1.3 Ontological Engineering

Ontological engineering is one of the methodologies

used to describe knowledge systematically. From the

viewpoint of knowledge base, “ontology is deﬁned

as a theory (system) of concepts/vocabulary used as

building blocks of an information processing system”

(Mizoguchi et al., 1995). Ontologies are classiﬁed

into two types according to the nature of the knowl-

edge described (Mizoguchi, 2003). One of them is re-

ferred to as domain ontology, which describes domain

knowledge, and the other is referred to as task ontol-

ogy, which describes knowledge about processes.

In the ontology development environment Hozo

†

each node represents a whole concept and has some

slots, each of which represents a part-concept. Each

whole concept consists of one or more part concepts

with part-of or attribute-of links (Figure 1). Hozo

supports the description of role concepts which repre-

sent a role that depends on the contents of each whole

concept. For example, a human being plays the role

†

http://www.hozo.jp/

Figure 1: Whole concept and part concepts.

of a teacher only in the context of school and not out-

side the school. In other words, every part concept in

the whole concept has a role to play within a given

context. In the context, a class of instances that can

play a role is deﬁned by a class constraint, and it is

called a role-holder (Kozaki et al., 2000).

Figure 2 shows the whole structure of a top-level

ontology YAMATO

‡

. According to YAMATO, an en-

tity is divided into three classes: physical, abstract,

and semi-abstract. While instances of physical class

need 3D space and time to exist, instances of an

abstract class need neither of them. Instances of a

semi-abstract class need only time to exist, and the

class contains a mind, representation, content, and

representation form. Representations such as nov-

els, poems, paintings, music, and symbols are dis-

tinguished from their proposition and form of repre-

sentation (Mizoguchi, 2004). A representation has a

content role played by a proposition and a form role

played by a representation form (Figure 3).

Although it is crucial in design tasks to manage

knowledge on impressions, no common method for

achieving this purpose has been established. In re-

lated research, a model of idea explanation styles

for a designer has been proposed; this model en-

ables designers to share their ideas about a new prod-

uct by adopting the ontological engineering approach

(Ogawaet al., 2009). The study did not directly model

vague ideas themselves, but modeled the explanation

style for these ideas. According to Ogawa’s frame-

work, we can practically model consumers’ impres-

sions of products, even though they are as vague as

ideas are.

In this paper, we introduce a framework to de-

scribe consumers’ impressions by adopting the onto-

logical engineering approach. As noted above, posi-

tive responses to irritations are derived from percep-

tual ﬂuency; therefore, we model perception and its

related concepts: awareness of an object and self-

report of impressions. We ﬁrst ontologically describe

awareness, perception, and self-report on the basis of

the Hozo and YAMATO environments to construct a

framework for sharing impressions. Then, by adapt-

ing our framework, we demonstrate an instance of im-

pression derivation during the task of appreciation of

Web pages that contains some designed elements.

‡

http://www.ei.sanken.osaka-u.ac.jp/hozo/onto library/

upperOnto.htm

PROPOSAL OF A FRAMEWORK TO SHARE KNOWLEDGE ON CONSUMER'S IMPRESSIONS

389

Figure 2: Whole structure of a top-level ontology YAM-

ATO.

Ex. Sentence

Ex. Natural language

Ex. Meaning

Figure 3: Slot structure of representation.

2 OUR FRAMEWORK FOR

IMPRESSIONS

2.1 Description of Awareness

Baruss deﬁned consciousness as “all subjective

awareness characterized by intentionality, and the ex-

plicit knowledge of one’s situation, mental states or

actions evidenced behaviorally” (Baruss, 1987). Sub-

jective awareness is referred to as subjective con-

sciousness, that is, “the stream of thoughts, feelings

and sensations that occur for a person”, and explicit

knowledge is referred to as behavioral consciousness

(Baruss, 2000). The latter is generated by the oper-

ationalization of the former. That is, behavioral con-

sciousness is deﬁned on the basis of operations that

infer other people’s consciousness during objective

studies.

Figure 4: State of being aware and awareness.

These two types of consciousness can be de-

scribed as a subclass of an external state deﬁned in

YAMATO (Figure 4). An of what role, which indi-

cates focus of attention, is played by an instance of

an object or a semi-abstract class, a thought, feel-

ing, or sensation class plays a referring to role, and

awareness is speciﬁed as a role-holder. YAMATO has

two kinds of propositions: representation-primary

and representation-secondary. Both representation-

primary and representation-secondary classes nec-

essarily depend on the representation that repre-

sents them (Figure 3). However, instances of the

representation-secondary class, such as facts, data,

and thought, indicate original content even before

they are represented. For example, a fact as an event

exists before human recognition, which allows the ex-

pression of the eventas a representation. In this sense,

the classes of feeling and sensation also belong to the

representation-secondary class.

2.2 Description of Perception

In YAMATO, the actor state action class which fo-

cuses on a change in the doer’s states, and the ob-

ject state action class, which focuses on a change in

ICAART 2011 - 3rd International Conference on Agents and Artificial Intelligence

390

the object’s states, are deﬁned. Perception can be re-

garded as the result of an action by which an object

is perceived; this result is reﬂected as a change in the

actor’s awareness. We deﬁned the class of perceive

as a subclass of the actor state action class. The ac-

tion to perceive needs a doer, an object, and a result

(Figure 5). The doer in the action to perceive has a

being aware state. Role-holders of an object role in

an action to perceive and an of what role in a state of

being aware belong to the same class. Relationships

among these concepts represent a situation where a

doer is aware of the presence of an object. The ac-

tion to perceive an object is regarded as an instance of

behavioral consciousness, which is described in the

previous subsection. Role-holders of a result role in

an action to perceive and a referring to role in a state

of being aware belong to the same class. This indi-

cates that the role-holder of the result role represents

the awareness of an actor. Thus, perception, which is

a result of an action by which an object is perceived,

is the instantiation of subjective consciousness, which

is described in the previous subsection.

Figure 5: Action to perceive and perception.

Since thought, feeling, and sensation are all sub-

classes of proposition, they require representation de-

pending on their contexts. However, any representa-

tion such as a mental representation that human being

forms in his or her brain is not included here. A pos-

sible deﬁnition of mental representation is a subjec-

tive phenomena, which is referring to thought, feel-

ing, and sensation. This deﬁnition can be regarded

as a specialization of the representation class. We

are concerned with the correspondence between in-

dividuals in the real world and the contents of con-

sciousness; hence, in our ontology, we do not include

contents represented in the brain, but include a self-

report as an output of the contents. In addition, we

can efﬁciently access the contents of consciousness

through verbal reports (Togawa, 2006), and the con-

tents of consciousness are indeed externalized as a

linguistic representation in the real world. Because

verbal reports focus on the description of the contents

of consciousness, we do not deﬁne mental represen-

tation here. However, we do not deny the existence of

mental representation.

2.3 Description of Self-report

We deﬁned an action called introspect as a special-

ization of the actor state action class. The introspect

does not have an object but has a result (Figure 6).

The role-holder of the result role of the action is de-

ﬁned as a self-report, and its class constraint is prim-

itive representation. A form role is played by an in-

stance of a word or a natural language class, and the

role-holders of a content role and the referring to role

belong to the same class.

Figure 6: Action to introspect and self-report.

3 DESCRIPTION METHOD FOR

WEB DESIGN IMPRESSIONS

3.1 Instances of Impression

Here, we can describe a good impression of a Web

page by adopting our framework. For example, an in-

stance where John has a positive feeling toward a Web

page is shown as Figure 7-a. In this ﬁgure, the im-

pression of the Web page is a positive feeling, which

is represented as an instance of perception. Figure 7-

b describes a scene where a Web page composed of

blue text and a highly contrasting white background

has the property of high ﬂuency. Generally, a Web

page contains some sentences as a component, and it

also has the attributes of text color and background

color. Perceptual ﬂuency, which emotionally gener-

ates a positive feeling in the perceiver, is represented

by the role concept of ﬂuency role-concept taking a

value of being high. After John experiences a positive

feeling, a subjective feeling is externalized as a self-

report. Figure 7-c shows a scene where John speaks

a sentence about his positive feeling.

PROPOSAL OF A FRAMEWORK TO SHARE KNOWLEDGE ON CONSUMER'S IMPRESSIONS

391

Figure 7: Instances of an impression.

3.2 Affections and Irritations

The description of an impression described above

help us to understand the correspondences between

affections and irritations. For example, a good feel-

ing described in the result of perceive is related to the

contrast between the text color and the background

color, which is referred to by the object of perceive in

Figure 7.

In our previous study, we proposed a model of

task knowledge in aesthetic appreciation (Muramatsu

et al., 2010). In this model, aesthetic appreciation is

divided into two tasks: representational differentia-

tion and emotional experience. These taskes are dis-

tinguished depending on what the perceiver is focus

on. However, the model ﬁnds it difﬁcult to estab-

lish correspondences between physical irritations and

the experience of beauty. In other words, beauty is

not a simple and direct consequence of physical irri-

tations. By Adopting our framework, we can describe

processes involvingaesthetic appreciationas an event,

and the correspondences can be resolved.

Figure 8: Example of grace event.

The concept of aesthetics includes multiple con-

cepts such as grace, sublimity, and ridiculousness.

When a perceiver emotionally experiences a positive

feeling during aesthetic appreciation, the experience

is identiﬁed as grace. Figure 8 shows an event where

a perceiverexperiencesgrace. This event has two pro-

cesses as part concepts. Both the processes are actions

to perceive but have different objects and results. The

ﬁrst process has an object that is composed of two col-

ored areas and a result that is referred to as harmony.

This indicates that combining colors result in a sen-

sation of harmony. The second process has an object

that is a result of the previous process and a result that

is referred to as grace. In this manner, we can resolve

the correspondences between physical irritations and

the experience of beauty.

4 CONCLUSIONS

In this paper, we introduced a framework for describ-

ing impressions by adopting the ontological engineer-

ing approach. We modeled the concepts of aware-

ICAART 2011 - 3rd International Conference on Agents and Artificial Intelligence

392

ness, perception, and self-report on the basis of the

top-level ontology. Awareness is deﬁned as a role-

holder of the referred to role, which is a part concept

of the being aware state. In a situation where a person

perceives an object, the perception is conceptualized

as a role-holder of the result role, which is a part con-

cept of an action to perceive. In such the action, a doer

is being aware state, and awareness and perception

have the same proposition. The self-report is deﬁned

as the representation of the proposition. In our frame-

work, we instantiate a case where a person has a good

impression of a Web page, and we describe the rela-

tionship between a perception and a stimulus in such

a case. This helps us to understand correspondences

between affections and physical irritations.

The current paper only presents an ontological de-

scription method for impressions. We need to study

more empirical cases related to impressions by adopt-

ing our framework. We believe that such empirical

studies would reveal the general rules governing the

correspondences between affections and physical ir-

ritations. In future work, we intend to expand our

ontological descriptions on the basis of the empirical

studies. Further research would help in clarifying the

fundamental nature of impressions.

REFERENCES

Baruss, I. (1987). Metanalysis of deﬁnitions of con-

sciousness. Imagination, Cognition and Personality,

6(4):321–329.

Baruss, I. (2000). Overview of consciousness research.

Informatica: An International Journal of Computing

and Informatics, 24(2):269–273.

Kozaki, K., Kitamura, Y., Ikeda, M., and Mizoguchi, R.

(2000). Development of an environment for building

ontologies which is based on a fundamental consid-

eration of ”relationship” and ”role”. In Proceedings

of the Sixth Paciﬁc Knowledge Acquisition Workshop,

pages 205–221.

Leder, H., Belke, B., Oeberst, A., and Augustin, D. (2004).

A model of aesthetic appreciation and aesthetic judg-

ments. British Journal of Psychology, 95(4):489–508.

Lindgaard, G., Fernandes, G., Dudek, C., and Brown, J.

(2006). Attention web designers: You have 50 mil-

liseconds to make a good ﬁrst impression! Behaviour

& Information Technology, 25(2):115–126.

Mizoguchi, R. (2003). Tutorial on ontological engineering

- part 1: Introduction to ontological engineering. New

Generation Computing, 21(4):365–384.

Mizoguchi, R. (2004). Tutorial on ontological engineering

- part 3: Advanced course of ontological engineering.

New Generation Computing, 22(2):193–220.

Mizoguchi, R., Vanwelkenhuysen, J., and Ikeda, M. (1995).

Task ontology for reuse of problem solving knowl-

edge. In Towards Very Large Knowledge Bases:

Knowledge Building and Knowledge Sharing, pages

46–59.

Muramatsu, K., Togawa, T., Kojima, K., and Matsui, T.

(2010). Content-oriented approach to knowledge de-

scription of aesthetic experience. In Proceedings of

the Kansei Engineering and Emotion Research Inter-

national Conference 2010, pages 856–865.

Ogawa, T., Nagai, Y., and Ikeda, M. (2009). An onto-

logical approach to designers’ idea explanation style:

Towards supporting the sharing of kansei-ideas in

textile design. Advanced Engineering Informatics,

23(2):157–164.

Reber, R., Schwarz, N., and Winkielman, P. (2004). Pro-

cessing ﬂuency and aesthetic pleasure: Is beauty in

the perceiver’s processing experience? Personality

and Social Psychology Review, 8(4):364–382.

Togawa, T. (2006). An approach to scientiﬁc understanding

of mind by deﬁning a set of all possible contents of

consciousness. Biocybernetics and Biomedical Engi-

neering, 26(1):5–20.

PROPOSAL OF A FRAMEWORK TO SHARE KNOWLEDGE ON CONSUMER'S IMPRESSIONS

393