氏名　Kumburehene Gamaralalage Deepika Tharangie
学位の種類　博士（工学）
学位記番号　博甲第553号
学位授与の日付　平成22年8月31日
学位論文題目　Kansei Apprising System to utilizeAffective Variability of Colours
in to Human Computer Interfaces　（色彩情報を用いたヒューマン・コンピュータ・インターフェースのための感性評価システム）
論文審査委員
　主査　准教授　アーシュ　マーラシンハ
　副査　教授　山田　耕一
　副査　教授　中村　和男
　副査　教授　福村　好美
　副査　教授　淺井　達雄

• 論文目次
• 論文要旨
• 論文審査の結果の要旨

［平成22（2010）年度博士論文題名一覧］ ［博士論文題名一覧］に戻る

1　Introduction　p.1
2　Background　p.3
　2.1　Motivation　p.4
　2.2　Aim and Objectives　p.4
　2.2.1　Aim　p.4
　2.2.2　Objectives　p.4
　2.3　Problem Definition　p.5
　2.4　Solutions　p.6
　2.5　Structure of the Thesis　p.7
　2.6　Summary　p.8
3　Literature Review　p.9
　3.1　Colour-Emotion　p.9
　3.2　Colour-Emotion Models　p.10
　3.3　Influence of Colour in Human Computer Interfaces　p.14
　3.4　Summary　p.15
4　Definitions of Technical Words　p.17
　4.1　Feeldings, Affect and emotions　p.17
　4.2　Affect　p.17
　4.3　Kansei　p.18
　4.4　Colour Harmony　p.19
　4.5　Semantic Differential Analysis　p.20
　4.6　Human Computer Interaction　p.21
　4.8　Summary　p.22
5　Background Theories　p.23
　5.1　What is Kansei engineering　p.23
　5.1.1　Working Principal of Kansei　p.23
　5.1.2　How Kansei is measured　p.23
　5.2　Colour Harmony Theories　p.24
　5.2.1　Chevreul's Colour Harmony Theories　p.25
　5.2.2　Ittens Colour Harmony Theory　p.26
　5.3　Abstract Circumplex Model of Emotions　p.26
　5.4　Summary　p.27
6　Approach:Kansei Colour Questionnaire　p.28
　6.1　Kansei Colour Questionnaire（KCQ）　p.28
　6.1.1　The User Groups　p.28
　6.1.2　Target Areas of the Kansei Colour Questionnaire　p.28
　6.1.3　Content of the Kansei Colour Questionnaire　p.29
　6.2　Kansei Colour Questionnaire Preparation　p.30
　6.2.1　Collecting kansei Words for kansei Colour Questionnaire　p.31
　6.2.2　Collecting Interfaces for Kansei Colour Questionnaire　p.32
　6.2.3　Selecting Basic Colour Palette　p.33
　6.3　Questionnaire Conducting Conditions and Procedure　p.34
　6.4　Questionnaire Data Analysis　p.35
　6.5　Kansei Colour Questionnaire Preparation　p.30
7　Analysis　p.37
　7.1　There are Individual differences in Affective variability of Colour　p.37
　7.2　Colour - Emotion Differences in Cross-Cultural Platform　p.38
　7.1.2　Age Specific Colour-Emotion Associations　p.42
　7.1.3　Gender Specific Colour-Emotion Associations　p.44
　7.1.3　Categorizing Colour-Emotion Associations drawn from Culture, Nature and Perceptual Beauty（Personal Differences）.　p.45
　7.1.4　How Coloour - Emotion relationship is made　p.51
　7.2　Influence of Affective Variability of Colours in Human Computer Interfaces.　p.52
　7.2.1　Tendency of Colours-Affective Associations　p.58
　7.2.2　Elaborating Affective Variability of Colours in Computer Interfaces Using Trait Kansei Words　p.60
　7.3　Affective Variability of Colour and Affective-Colour Space　p.67
　7.3.1　Colour-Affective Space　p.70
　7.3.1　Colour-affective Space in Abstract Circumplex Model　p.73
　7.4　Usefulness of Conducting Kansei Surveys in Designing Human Computer Interfaces　p.75
　7.5　Summary　p.76
8　Design:Kansei Aapprising System　p.78
　8.1　Kansei Appraising System Model　p.78
　8.1.1　Affective space　p.80
　8.1.2　Product space　p.80
　8.1.3　Measuring phase　p.80
　8.1.4　Analysing Phase　p.81
　8.2　Making Three Colour Combinations for On-line Kansei Apprising System　p.82
　8.3　Selecting Kansei words for On-line Kansei Apprising System　p.82
　8.4　System Performance　p.83
　8.4.1　Advantages of the system　p.84
　8.4.2　Disadvantages of the System　p.85
　8.5.1　User Interface of the Online System　p.86
　8.5.2　System Architecture of the Online Kansei Appraising System　p.87
　8.5.3　Technology Used for implementing Online Kansei Apprasing System　p.89
　8.5.4　Reports Produced by the Online Kansei Apprasing System　p.91
　8.6　Validating the Kansei Appraising Model　p.91
　Age-wise Comparison　p.93
　Gender-wise Comparison　p.93
　Status-wise Comparison　p.94
　Culture-wise Comparison　p.94
　8.6　Summary　p.102
9　Conclusion　p.103
　9.1.1　Problems encountered　p.103
　9.1.2　Limitation of the Research　p.104
　9.1.3　Achievement of Research Objectives　p.105
　9.2　Conclusion　p.106
　9.3　Future work　p.107
References　p.109
Authors Publication List　p.ii
Appendix 1: Information Collection　p.iv
Appendix 1I: Colour Selection　p.v
Appendix 1II : Adjective Selection　p.vi
Appendix IV : Testing with Bi-polar Scales　p.vii
Appendix V : Class Diagram of Online Kansei Appraising System　p.viii
Appendix VI : Sequence Diagram of Online Kansei Appraising System　p.ix
Appendix VII : Tree Classification Rules　p.x
Appendix VIII : Reports　p.xv

　Colour is an imperative aspect of our visual experiences. If colours are used properly, they can be powerful tools in improving the usefulness of computer interfaces in a range of areas. Colour combinations that are used in the computer interface components such as background, banners, graphics, etc. influence the users’ subsequent opinion and satisfaction in both visual and usability aspects. Therefore as an attempt to provide a better design solution, designers’ resort to the aid of a numerous existing user interface guidelines, colour models, formulae, and software solutions etc. However, even after consulting all these existing systems, designers still fail to cater for the fullest satisfaction of users. In searching for the possible reasons to this inability, a cross-cultural kansei survey was conducted. The research learnt that undermining the fact of the affective variability of colours by designers as a prominent factor that led to the less user-friendly user interfaces. More than 150 Sri Lankan and Japanese participant groups took part in the survey, and following hypothesis were tested. There are individual differences in affective variability of colours and affective variability of colours influence human computer interaction. The result was statistically significant at more than 95% level of confidence, confirming the existence and influence of affective variability of colours and individual differences in computer interface designing. Therefore, to examine the problem at hand in depth, a literary survey was carried out on existing colour-emotion association systems. As a result, following drawbacks were identified. Firstly, most of the existing systems are abstract and generalised approaches rather than customizable personalized approaches. Secondly, most of these systems rely on already made databases or indexing systems to select appropriate colour solutions. Thirdly most of the already existing systems were particularly designed for specific subject domains （Textiles, Fashion, Architecture, Product designing, etc）, and rarely considered the colour-emotion associations in human computer interfaces. Finally, although there are practical approaches as harmonious colour selecting tools, these software do not consider the affective variability aspect of colours yet. Hence, to plug the drawbacks of these already existing systems and findings, the kansei survey laid out the necessity for a personalized practical approach, which will incorporate the affective variability of colours while providing harmonious colour schemes. As an attempt to fulfill this requirement in demand, this research proposes a new research paradigm and a solution methodology to appraise prospective users’ affective variability of colours for given harmonious colour schemes, in return to provide better colour design solutions. As the result, kansei appraising system model was proposed. Finally, to validate the kansei appraising model, online kansei appraising system was implemented. Online kansei appraising system incorporates kansei engineering principles with colour harmony theories to gain and analyse users’ affective variability towards colours. The kansei appraising system is a customizable practical approach, which provides an interface with the necessary middle layer application that enables designer / engineer to conduct surveys, analyse data, identify the personal trends and patterns in their selected sample groups, and make appropriate and specialized decisions towards user satisfaction. To explain the working principal of kansei appraising model, a survey was conducted for 150 participants in both Sri Lanka and Japan employing the online kansei appraising system. Result was elaborated using Decision Tree procedure. Decision Tree was utilized to create a tree-based classification model of the results to maximize the insight into the dataset, Identify entities, which are likely to be members of a particular group and uncover the underlying structure of the dataset to validate the kansei appraisal model. The outcome of the evaluation was notable and biasing with the proposed kansei appraising model. Furthermore the results were overwhelmingly suggested the congruity of the study. The elaborated tree structure, generated from the result reveals that there are wide array of differences in colour?emotion association. These differences can be recognized as individual differences within groups （groups: culture, age, gender, status） and between groups. Further, this complex structure generated from the system results affirms the validity of the kansei appraising model and usability of kansei appraising system to analyse, compare, contrast and synthesis each an individual substantial factors in the affective variability of colour in a given user group. In addition to the fulfillment of the research aim （kansei appraising system model）, objectives of the research also were met. As achievements of project goals, this research contributed six more deliverables to the existing research field. Namely, proposed a model for how kansei works, proposed a model for working principal of kansei engineering, designed and implemented the kansei words database, development of colour combination database and bi-polar scale database, and design and development of an emotion-colour survey method （kansei colour questionnaire）. As the last and eminent deliverable, the online colour appraising system was implemented. Due to lack of resources and facilities, research had to face many hindrances. However online kansei appraising system was able to overcome many of these limitations. As a future extension to this research work, following areas are considered. Kansei appraising system consists of an affective space and a product space. Product space consists of product properties and an affective space of affective/kansei words. Although there is a mechanism to select kansei words, there is no standardized mechanism to select product properties. Therefore as a future development, the research aims at proposing a new standard methodology to identify product properties in any subject domain. Although this research domain is particularly limited to colour and computer Interfaces, the proposed kansei appraising model can be utilized for any subject domain by simply filling the affective space with kansei words in the selected domain and product space by attributes or properties from the selected domain.

　本論文は、「Kansei Apprising System to utilize Affective Variability of Colours in to Human Computer Interfaces（色彩情報を用いたヒューマン・コンピュータ・インタフェースのための感性評価システム）」と題し、

　The structure of this thesis is organized as fallows.
　Chapter 1 introduces the research topic to the reader and sets the scene and problem statement.
In the 2 chapter, wider appreciation of the topic is provided along with motivation, research aim and objectives of the research, the problem st　atement and solutions.
　Chapter 3 presents a brief review of literature relevant to this thesis.
　Chapter 4 briefly introduces the new terminology used in the thesis.
　Chapter 5 presents the background theory incorporated in the research.
　Chapter 6 presents how theories have been incorporated in the research.
　Chapter 7 forwards the hypothesis, procedure of hypothesis testing with the critical analysis of the results comparison with existing research in the research domain.
　Chapter 8 puts forward the proposed model for Kansei appraisal system to utilize affective variability of colour into human computer interfaces and further to validate the designed model, practical application of the proposed model, online Kansei appraising system is discussed.
　Finally, the problems encounterd, limitations, conclusion and future research works are described in chapter 9.

　よって、本文論は工学上及び工業に貢献するところが大きく、博士の学位論文として十分な価値を有するものと認める。