Culturicon Design Model for Social Mobile Application

The usage of emoticon in computer-mediated communication has been growing rapidly among users, especially in social media. Emoticon has been used to express feelings, emotions, gestures, actions and places. Despite the growing number of emoticon users around the world, study on the cultural elements of the emoticon is still lacking. This research aims to propose a model for the development of Culturicon, which is Cultural-Based Emoticon. In doing so, a verification process must be done to the proposed model to ensure that the model is well verified. Expert review method was used for the verification method. Experts from the field of Human-Computer Interaction, User Experience and cultural study especially the academicians were chosen. In addition, application developer and graphic designer also were chosen as expert from the industry. The experts were approached by email and performed the verification by answering online questionnaire provided. The result obtained from these experts were analyzed and amendment were made based from the comments and suggestions. Results showed that 91% experts agreed the connections and flows of all components in the proposed model are logical and readable. Expert verification is important to ensure that the model is being develop correctly. By having this model, it can aid designer and developer in designing meaningful and effective culturicon. Keywords—Emoticon, culture, expert review, human-computer interaction.


Introduction
Mobile messaging applications such as WhatsApp, Telegram, Facebook Messenger, Twitter and WeChat provide emoticon selection to be used for Computer-Mediated Communication (CMC). The usage of emoticon has evolved from time to time. At first, the usage of emoticon is created by Scott E. Fahlman in the form of ASCII symbols by using keyboard characters [1]. Then, the emoticon has evolved to the form of pictogram, such as the emoji and sticker [2]. Previous research has divided the category of emoticon into three categories: 1) ASCII symbols, which are the emoticon made from keyboard character. 2) Preload Static Images, which are the cartoon images design by professionals.
3) Preload Dynamic Images, which are the short video or animation. 4) Local Upload Images, which are the images uploaded by the user [3].
These emoticons can show various expressions, from facial expressions to anything that relate to our daily lives such as cultural element.
The creation of emoticon brings a lot of benefit to the user for their communication in CMC because it helps the user to better express their emotion by providing better additional cues [4], strengthening the intensity of the message [5] and improve the efficiency of communication [6], [7]. These make the usage of emoticon become more popular day by day among heavy users of the mobile messaging application. Despite all these benefits, research has shown that emoticon also lacks in the terms of emoticon selection, emoticon misinterpretation and misused especially with people from different culture [8]- [10]. Due to this lacking, a Cultural-Based Emoticon (Culturicon) Design Model has been proposed to assist designer in designing appropriate Culturicon especially for Asian countries. The proposed model needs to be verified to ensure the development of the model is correct. Expert review method has been chosen for the verification process.

Related Work
Cultural models from other disciplines have been applied in Human-Computer Interaction (HCI) research. Among the cultural models that have been applied in HCI research are Hofstede Culture Model and Trompenaars Culture Model [11]. Research by Gould, Zakaria and Yusof used dimension of Power Distance and Individualism-Collectivism from Hofstede Culture Model and a dimension of Specific Relationship from Trompenaars Culture Model to compare representative's websites [12]. These dimensions were used to structure the differences between the cultures in Malaysia and United States during the usage of websites. They found these 2 cultures own their different ways of interacting and priorities.
Meanwhile, Callahan studied university's websites from 8 different countries using Hofstede Culture Model [13]. He compared the differences between cultures on the layout design, type and frequency of images and number of links per page of the university's websites. They found that there are correlations of logo image usage with dimension of high-Power Distance and figurative images usage with dimension of Masculinity. Dimension of Masculinity also was used in research by Dormann to overlook on the different emotions and values expressed on university website from three countries [14]. She found that the feminine countries expressed stronger feminine values as described in Hofstede Culture Model. Based from the related work, this research aims to apply the cultural models in designing emoticon to be used in CMC.

Model Development
Based on the proposed initial model by [15], further study has been conducted to determine the suitable cultural dimension to represent Asian countries. In addition to that, characteristics of the emoticon created based on the model also were determined. As a result, the proposed model will come in 2 phases, Cultural Dimension (Phase 1) followed by HCI Icon Principle (Phase 2).

Cultural dimension
For Cultural Dimension, there are 3 cultural models considered in this research. They are Hall Culture Model [16], Trompenaars Culture Model [17], and Hofstede Culture Model [18]. After comparing these 3 models, it is found that some dimension from these 3 models are similar. Since the Hofstede Culture Model is the most popular among the other 2 models [11], researcher choose to use the naming of dimension from the Hofstede Culture Model. Hofstede Culture Model is popular because of its comprehensive framework, empirically confirmed and heavily cited. Table 1 below describes the cultural dimensions along with the trait and criteria.  Table 1 shows Phase 1 of Culturicon Design Model which is the cultural dimension. Phase 1 consists of cultural dimension, trait and criteria. The level of the selected cultural dimensions are also stated which are high power distance [12], [18]- [20], high collectivism [12], [19]- [21], low uncertainty avoidance [12], [19], [20], moderate masculinity/femininity [12], [19] and long-term relationship [12], [19], [20]. In each of the dimensions, there are traits as the classification for the dimensions such as national, social norm, religion, political, work, school and family. In each trait is the criteria that suggest the element for the cultural dimension.

Human-computer interaction icon principle
The next part is Phase 2, which is the HCI Icon Principle. After determining the cultural element based on Phase 1, the next step is to design emoticon that fulfills the principle provided in the Phase 2. Table 2 depicts HCI icon principle and its criteria.

Methodology
In verifying the culturicon design model, expert review method was chosen. The verification of the model was done by the experts from the related domain. The reason of expert review is chosen for the verification process is because expert review is fast, cheap, intuitive and easy to motivate people to do it and does not require advance planning [22], [23].

Expert selection
In order to appoint the experts, a letter of nomination was sent to the chosen experts through email to seek their acceptance to participate in the verification process. The verification process for Culturicon Design Model involved 4 HCI experts, 2 culture experts, 2 arts and design experts, 1 application developer and 2 graphic designers from the industry, that make up 11 experts in total. The 11 experts chosen is sufficient as Nielsen and Molich suggest 3 to 5 experts to perform expert review [22]. The HCI, culture and arts and graphic design experts are academicians who are the faculty member of institutes of higher learning and practitioner with more than five years of experience. Meanwhile, the application developer and graphic designer experts come from the industry whom are also the expert that have at least 5 years of experience, as suggested by Rogers and Lopez [24]. Table 3 describes the demographic data of the selected experts.

Verification process
After the chosen experts agreed to participate in this verification process, another set of email were sent to them that contain the introduction about the research conducted, the instructions on doing the verification process, and the proposed model itself along with the online verification form link to answer questionnaire for verification process. The experts then reviewed and verified the model through the online questionnaire provided. After the expert reviewed and verified the model, the results and feedback were analysed, and amendment were made.

Verification result
There are 3 parts of verification that need to be answered by the experts. The first part is about the verification of the overall model, the second part is the verification on cultural dimension, and the third part is the verification on the HCI icon principles.   Table 5 shows that 9 experts (82%) agree that High Power Distance is easy to understand while 2 experts (18%) indicated that the component need some explanation. For High Collectivism, 6 experts (55%) agree that the component is easy to understand, while 5 experts (45%) stated that it need some explanation. The result for Low Uncertainty Avoidance and Moderate Masculinity/Femininity are similar where 6 experts (55%) agree that the component easy to understand, 4 experts (36%) indicated that it needs some explanation and 1 expert (9%) stated it needs very detail explanation. For the last component, Long-Term Relationship, 5 experts (45%) agree the component is easy to understand, 4 experts (36%) indicated that it needs some explanation and 2 experts (18%) stated that it needs very detail explanation. Table 6 shows the verification result on HCI Icon Principle.  Table 6, for the principle of Familiar, Understandable, Memorable, 9 experts (82%) agree that it is easy to understand, while 2 experts (18%) indicated it needs some explanation. For the component of Attractive, Informative and Compact, 10 experts (91%) agree that it is easy to understand, while 1 expert (9%) indicated it needs some explanation. For component of Distinct and Unambiguous, 8 experts (73%) agree that it is easy to understand while 1 expert (9%) indicated it needs some explanation and 2 experts (18%) indicated it needs very detail explanation. For component of Coherent, 6 experts (55%) agree that it is easy to understand while 5 experts (45%) indicated that it needs some explanation. For component of Legible, 7 experts (64%) agree that it is easy to understand while 4 experts (36%) indicated that it needs some explanation. Lastly, component of Extensible, 9 experts (82%) agree that it is easy to understand, while 1 expert (9%) indicated that it needs some explanation and 1 expert (1%) indicated that it needs very detail explanation.

As shown in
As a part of verification process, experts may provide comments and feedback based on their perspective regarding the proposed model. Figure 1 and 2illustrate the feedback result on the overall proposed model.  Figure 1 shows the result regarding the connection and flow of the model. 10 experts (90%) agree that the connections and flows of the model are logical, while 1 expert (9%) disagree. Figure 2 shows the result regarding the readability of the model. 10 experts (90%) agree that the model is readable, while 1 expert (9%) disagree.
For the last section, experts were asked to give any comment and suggestion regarding the proposed model. From the 11 experts, only 6 experts gave comments and suggestions as presented in Table 7.

E1
Stated that there are too many components provided in the model and ask whether all of these components are needed in developing Culturicon. Need to enlarge the font size. E5 Scope of culture is broad. The model needs to be more focused and precise towards the culture it wants to express.

E11
Some components need explanations as they are hard to understand and to interpret into graphical element.

Connection and Flow of the model E1
Need to consider and identify the mandatory and optional components of the model. Provide simple OD or legend for each component, provide relationship between components to show clear connection.

E9
Should provide process flow.

E6
Overall are good and interesting E7 At this stage, description explained in words and it will be interesting when designer begin to sketch the icon representing the cultural-based.
Agreed with the proposed model and will be interesting to see the findings.
As shown in Table 7, there are comments and suggestions regarding the components and the connection and flow of the model. For comment on the component in the model,

91% 9%
Overall, the proposed model is readable

Yes No
E1 stated that there are too many components provided in the model, especially in the HCI Icon Principles where there are 11 principles provided. E1 suggested that some component need to be reviewed again whether they are needed in the model or not.
E5 stated that the actual cultural scope is broad to be fitted in the model. The model needs to be more focused and precise towards the culture it wants to express. The criteria need to have more detail on the culture of Asian countries so that the model really depicts its culture that make user easy to understand and use it as the guideline to create Culturicon. E11 commented that some components are difficult to understand due to lack of explanation, especially in the Cultural Dimensions part. E11 stated that some people are not familiar with the term used for the Cultural Dimensions.
For the flow of the model, E1 suggested that the model shows which component is mandatory and optional to ensure the flow of the model is clear. This is because, when user wants to use the model, user needs to choose either one or many components in the Cultural Dimensions, while for HCI Icon Principle, user need to fulfil all the principles. To overcome this, E1 suggested that the model provides legend for each component and relationship between components. E1 also stated that this model is significant in creating the Culturicon. E9 also stated give suggestion that there should be a process flow so that it will be easier for the user to understand how the model works. Lastly, for the overall comment on the model, E6 and E7 agreed that the model is good and interesting. E7 also agreed with the model and interested to see the finding of this research. The comments and suggestions received from these experts are taken into consideration for the modification of the model. These comments and suggestions were examined carefully, and the model were modified accordingly.

Result and Discussion
Based on the verification result, amendments were made in line with the experts' comments and suggestions. Figure 3 below shows the revised version of Culturicon Design Model. The revised model has incorporated a process flow that makes it easier for user to follow, based on the suggestion from E1 and E9. By having the process flow, it helps user to acknowledge on how to use the model, where should they start first and what is the next step. The process flow starts with Phase 1 which is Identify Cultural Element. This phase is where user will choose either one or many cultural elements from the cultural dimensions provided. At least one cultural element must be selected. Next phase is Design Culturicon. For this phase, user need to fulfill all the HCI icon principles provided when designing Culturicon. There are 8 principles provided after amendment has been made. The design should take consideration from all 8 principles provided in the model. Next phase is Produce Culturicon. After user design the culturicon, they need to produce the culturicon based from the chosen cultural element and the design that fulfilled HCI Icon principles.
For the detail Culturicon Design Model, in Phase 1, there are changes for the traits and criteria that were made based from the suggestion by E5. These changes were made in order to accommodate the cultural perspective of certain Asian countries. For High Power Distance, 82% responded agree it is easy to understand, while 18% responded it need some explanation. Changes were made for Political and Work traits by adding additional information which is traditional/modern. This is to give an overview to user whether to design it in the context of traditional or modern element. For High Collectivism, 55% responded it is easy to understand, while 45% responded it need some explanation. Changes were made to the trait of National, Social Norm, Work and Family. In National trait, the criteria of extended family and tribe were moved to trait of Family. Social Norm trait were added to add criteria of obedient to custom and rule and value interpersonal harmony. In trait of Work, the previous criteria were replaced with work in group and support each other and maintain harmony environment.
For Low Uncertainty Avoidance, 55% respond easy to understand, 36% responded it need some explanation and 9% responded it needs very detail explanation. Changes made for trait of Political, Religion and Work. In trait of Political, criteria citizen protest are accepted and fewer rules and regulation were replaced with government changes are accepted. In trait Religion, criteria little persecution to belief is removed. Trait of Work is added with the criteria less loyalty and at ease with changing job. For Long-Term Relationship, the trait Social Norm, Work and Family were added. The previous criteria were sorted accordingly. Criteria be thrifty is added to trait of Family. All the changes and amendments are made based from the suggestion by the E5, which is the cultural expert For Phase 2, changes made were: the omission of HCI icon principle of Unambiguous, Extensible and Compact based on the suggestion by E1. The omission of principle Unambiguous is because its criteria are similar with the principle Distinct. The principle Extensible was omitted because its meaning and criteria are similar with principle Informative and Legible. While principle Compact was omitted because it is not a critical principle in designing Culturicon. Table 9 below shows the details of the Culturicon Design Model.

Conclusion and Future Work
Based from the suggestions and feedback from the expert, amendment have been made. This is to ensure that the model is being develop correctly. As this model contain cultural element, advice from the cultural expert was used as the guidance during the amendment process so that the cultural element can be suitable and appropriate with Asian countries as its target user. By correctly develop the proposed model, it can assist user, which is the designer in designing meaningful and effective Culturicon to be used by the mobile messaging user. After the proposed model has been verified through the expert review, next step is to perform the validation process. The verification process is to ensure the development of the model is right, while the validation process is to ensure the right model is developed. To perform the validation process, the target user, who is the designer needs to use this model to design the Culturicon. Then, after designer finished designing the sample Culturicon, the sample Culturicon will be used by the mobile messaging user in their mobile messaging application for their daily communication within a given time frame. A focus group session will be held once the given time frame ended to discuss the usefulness of the Culturicon developed.

9 Authors
Mohd Zhafri Mohd Zukhi is a lecturer at Universiti Teknologi MARA, Kedah. He interested on Human and Mobile interface design research. Email: zhafrimohd23@gmail.com Azham Hussain is the Associate Professor of Software Engineering at UUM School of Computing. He is the founder and head of Human-Centered Computing Research Group which is affiliated with the Software Technology Research Platform Center at School of Computing, Universiti Utara Malaysia, 06010, Kedah, Malaysia. Assoc. Prof. Azham Hussain is a member of the US-based Institute of Electrical and Electronic Engineers (IEEE), and actively involved in both IEEE Communications and IEEE Computer societies. Azham is published in the areas of software evaluation and testing, user behaviours, group collaboration, and ubiquitous and mobile technology design. He has authored and co-authored more than 100-refereed technical publications, served as reviewer and referee for refereed journals and conferences on computing as well as the examiner for more than twenty doctoral and postgraduate scholars in his research areas. Email: azham.h@uum.edu.my Husniza Husni is a senior lecturer at School of Computing, Universiti Utara Malaysia, 06010, Kedah, Malaysia and currently serves as Coordinator for Computing Professional Enrichment and Development Division (CoPED). She holds a Bachelor degree majoring in Artificial Intelligence (with Honours) from Universiti Utara Malaysia in 2002 and a Master degree in Computer Science from the University of Western Australia in 2005. Later, she received her her Ph.D. (Information Technology) from Universiti Utara Malaysia in 2010 and her doctoral work involves specific modeling of lexical model for dyslexic children's reading in Malay for the purpose of automatic speech recognition (ASR). She uses the major theory of dyslexia, the phonological deficit theory, to serve as fundamental in developing an ASR model specifically for dyslexic children's reading with attention given to the common reading patterns. Her research interest continue to grow in the realm of automatic reading tutor and interactive literacy education for useful educational technology to assist children with reading difficulties. Email: husniza@uum.edu.my