Keywords

1 Introduction and Background

There is a growing acknowledgement in the Web community that the user’s cultural background is, to some extent, essential and need to be taken into consideration during the design process of localized websites (i.e., websites adapted for a specific country, region, society, …) [1–4]. Therefore, we have done several studies aiming at verifying and investigating the relationship between features of the website and the users’ cultural background in terms of web usability [5–8]. Our experiments showed that it is not possible to have an absolute and clear-cut set of cultural markers to be used for designing localized websites. In addition, one single set of anthropological cultural dimensions could, in fact, be a poor choice because different levels of localization may be needed in different situations. Therefore, 5 different levels of cultural markers organized as a pyramid has been proposed [8]: (1) the e-culture level – for non-localized (thus rather international) websites, (2) the settled cultural level – for semi-localized websites, (3) the broad cultural level – for localized websites, (4) the variable cultural level – for highly localized websites, and (5) the vista cultural level – for fully cultural localized websites. For each level of localization, a group of cultural markers for a set of website design elements is provided, as well as a specific number of anthropological cultural dimensions to be considered for a specific localization level.

The website design elements considered for the different levels are: (1) Text on websites; (2) Layout and Organization; (3) Colors; (4) Pictures, Graphic Elements, and Sound; (5) Interaction; and (6) Navigation. The anthropological cultural dimensions considered are 16 existing cultural dimensions: Human Nature Orientation, Individualism vs. Collectivism, Internal vs. External Control, Time Orientation, Authority Conception, Context, Gender Roles, Power Distance, Uncertainty Avoidance, Universalism vs. Particularism, Achievement vs. Ascription, Affective vs. Neutral, Specific vs. Diffuse, Experience of Technology, Face-Saving, and International Trade and Communication.

In [9] we described the Cultural Conceptual Model (C2M) (expressed in Object Role Modelling (ORM) [10]) based on the Cultural Markers Pyramid. And did transformation of C2M to Ontology (OWL [11]) and called it Localization Ontology for a more practical oriented format [12].

In this paper, we describe how the Cultural Conceptual Model (C2M) is used to develop the Localized Website Design Advisor tool. The tool uses the Localization Ontology and a Cultural Markers Knowledge Base extracted before and explained in [9].

The paper is structured as follows. In Sect. 2, we explain the Localized Website Design Advisor (LWDA) and why such a tool is needed. We then explain in Sect. 3 the technical details of LWDA. In Sect. ‎4, we describe how to use the Localized Website Design Advisor in practice. Conclusions are drawn in Sect. 5.

2 Localized Website Design Advisor (LWDA)

LWDA is an acronym for Localized Website Design Advisor. It is a web-based tool that uses the Localization Ontology [9] as well as a knowledge base, called the Cultural Markers Knowledge Base, which contains information and cultural markers specifications for different target societies.

LWDA tool is not a website design tool; rather it is an advisor-tool that provides guidelines that developers should take into consideration when designing a localized website for a given localization level, society, language, main website domain, and some website related domains. This means that the main input for the tool is: (1) Localization level that one wants to achieve “1 to 5”, (2) Target society “Country”, (3) Website language, (4) Main website domain, and (5) Related website domains.

2.1 Necessity of a Localized Website Design Advisor Tool

Knowledge about the localization of websites is scattered over many different (mostly academic) resources. This impedes the proper development of localized web sites in practice. Website developers tend to rely on their common sense and on what others did to localize a website. However, such an approach is not always resulting in an appropriate solution, as what works in one context may not work in another context. Therefore, LWDA was developed to provide a central repository for knowledge about designing localized websites. To make the repository easy to use in practice, it was designed as an advisor tool that gives to-the-point guidelines to web developers for localizing a website in a specific context. When specific information is not yet available in the repository, the tool can be used (by experts) to fill in this information. Furthermore, LWDA can also be employed for the evaluation of the localization of existing websites by using the guidelines to verify whether the website satisfies the guidelines.

3 Localized Website Design Advisor (LWDA) Technical Details

LWDA was developed using ASP.NET (Microsoft web application framework) [13] and AJAX (Asynchronous JavaScript and XML) [14] is used for the communication between the browser and server, which removed the need for a whole page to be reloaded after each interaction.

In order to communicate with the ontology, LWDA makes use of XSLT [15], to dynamically generate a questionnaire based on XML documents. Furthermore, LWDA is connected to a Microsoft access database, which is used for the cultural knowledge base to store information about the cultural markers of target societies.

3.1 LWDA Architecture

As illustrated in Fig.‎ 1, the LWDA application is subdivided into six processes. This section describes these processes, as well as the internal relationship between them.

Fig. 1.
figure 1

Localized Website Design Advisor (LWDA) Architecture

Full size image

LWDA contains two information repositories: (1) the Localization Ontology, and (2) the Cultural Markers Knowledge Base.

  • The Localization Ontology is used to provide a common and unambiguous understanding of the website design elements and cultural dimensions and their interdependencies for the different localization levels. It actually contains all information provided by the Cultural Markers Pyramid.

  • The Cultural Markers Knowledge Base contains information and cultural markers information about specific societies and specific website domains.

We separate between those two information repositories for the following reasons:

  1. 1.

    The Localization Ontology is used for the abstract guidelines as specified in the Cultural Markers Pyramid. The Localization Ontology should only be changed by a researcher specialized in culture. For example, somebody who can decide whether or not the “number of links in a group” is a cultural marker for a specific localization level. This kind of information should come from research and long-lasting investigations.

  2. 2.

    The Cultural Markers Knowledge Base is a information repository containing information about (1) a particular country, (2) a specific website domains, and (3) a specific localization level. The stored information can be considered as an instantiation of the abstract guidelines provided by the Localization Ontology. The Cultural Markers Knowledge Base (or parts of it) can be maintained by somebody who is a culture expert for a particular target culture.

  3. 3.

    The social as well as the digital culture are changing constantly [8, 16]. Therefore, information in the Cultural Makers Knowledge Base may change over time. For example, maybe today the maximum number of links in a group has to be 7 for Chinese people in a Health website, and maybe after a while this needs to be changed to 8. As long as the number of links is still a cultural marker, the value can be updated in the Cultural Makers Knowledge Base. If the number of links would no longer be a cultural marker, then the Localization Ontology needs to be updated (separation between the marker itself and the value of the marker).

  4. 4.

    Having the cultural markers characteristics in a knowledge base makes it easy and flexible for external application to use this data.

In order to provide a better understanding of LWDA, the following section describes the different processes of LWDA, as well as the roles of the two cultural repositories.

3.2 LWDA Processes

Here we describe the LWDA processes using a real example. Suppose we need to design a localized website, then (as shown in Fig. ‎1), the following processes are involved:

  1. 1.

    Localization specifications request:

    In this first step, the website developer needs to give some information about the website he or she wants to design. Suppose that a website developer wants to design a health website for China, wants to achieve the localization level 4 (Variable), and the website language is Chinese. After providing this information to LWDA, the tool takes over the process and passes to the second process.

  2. 2.

    Checking the availability of localization specifications:

    In this process, the tool checks the Cultural Markers Knowledge Base trying to find out whether the knowledge base contains information about such a request.

Suppose that there is no such information yet available then the tool passes to process 3.

  1. 3.

    Create a new project:

    The LWDA tool contacts the Localization Ontology asking for cultural markers specifications for the information entered by the website developer.

For example, the ontology generated the following specifications: “VariableNavigation contains at most NavigationDepth”; “NavigationDepth is accepted by TargetCultureGroup” . This means that, in the variable level, the navigation depth is important to take care of and it is important to identify the depth of website navigation as accepted by the target culture.

The information collected from the “Localization Ontology” is still abstract and no values are assigned (e.g., the actual value of NavigationDepth is not provided). For that, a questionnaire is generated from the Localization Ontology to be given to social and digital cultural experts/researchers to fill-in. This is explained in the following process, process 4.

  1. 4.

    Cultural questionnaire:

    In this process, all the specifications obtained from the Localization Ontology are formulated in the form of a questionnaire. The generated questionnaire is easy for non-computer specialists to deal with. For the example, the questionnaire will contain a question on the maximal navigational depth accepted by Chinese people for Health websites. The answer to this question is expected to be provided by social and digital cultural experts/researchers who have deep knowledge about the target country China.

After the questionnaire has been filled in, the answers are processed by the tool. The tool ensures that the mandatory questions in the questionnaire are answered. Next, the information is processed and stored (process 5).

  1. 5.

    Processing and store information:

    The tool will pass the answers on the questioner to the Cultural Markers Knowledge Base to store it. In this way, the information can be reused for later projects (e.g., somebody needs to have website localization specifications for a similar website). But it could also be used for querying the knowledge base by other applications.

The next process is the process with number 6.

  1. 6.

    Processing and deliver localization specifications:

    Based on the information in the Culture Markers Knowledge Base, the tool generating a report contains the requested localization specification guidelines.

Note that when the requested information was already in the Culture Markers Knowledge Base at the time of the request (process 2), processes 3 to 5 are skipped.

4 Using LWDA

In this section, we illustrate the use of LWDA in practice. For example, suppose we need to design a localized website for China in the domain of Health. Also suppose that the relevant information is not yet in the Culture Markers Knowledge Base.

First, the website developer provides LWDA with the required information about the target website. This is entered at the start page (Fig. ‎2).

Fig. 2.
figure 2

LWDA start page

Full size image

As an example, suppose the user has entered the following information:

  • Website name “Project name”: The Public Health Portal of China

  • Localization level required: level 3, Broad

  • Target society “Country”: China

  • Language: Chinese

  • Website is belongs to one Website Domain: Health

  • Website domain related to: Government and News

Next, the system checks if the Cultural Markers Knowledge Base already contains the requested information. Remember that we suppose that this is not the case. Therefore, the LWDA tool will generate a “Cultural questionnaire” from the information in the Localization Ontology. This questionnaire should be filled-in by a social or cultural expert and needs to be returned to LWDA for processing. Figure ‎3 shows the structure of the generated questionnaire for the target website.

Fig. 3.
figure 3

Generated cultural questionnaire

Full size image

The questionnaire is divided into two main categories. The first category is for collecting information about website design elements, and the second category is to collect information about anthropological cultural dimensions.

Figure ‎4 shows an example of collecting information about one of the website design elements “Colors”. The questionnaire provides information and raises questions about colors. Figure 4 shows that the expert gave three notes about colors; these notes are cultural markers values.

Fig. 4.
figure 4

Generated cultural questionnaire: question about colors

Full size image

The following figure, Fig. ‎5, shows examples of the anthropological cultural dimension questions. The expert has added notes in the “ExpertNote” field.

Fig. 5.
figure 5

Generated cultural questionnaire: question about the gender roles anthropological cultural dimension

Full size image

After answering the questions, the expert should return the questionnaire to the LWDA tool for processing. Afterwards, LWDA stores the information for later reuse and uses it to automatically generate specific guidelines for the target web-site.

The following snapshots: Figs. ‎6, ‎7 and 8, are a part of the specification report generated. The generated guidelines are in a pseudo natural language (English).

Fig. 6.
figure 6

Target localized website specification guidelines

Full size image
Fig. 7.
figure 7

Target localized website specification guidelines for interaction

Full size image
Fig. 8.
figure 8

Target localized website guidelines for uncertainty avoidance and gender roles

Full size image

Figure ‎7 shows a part of the specification report for “Interactions”.

Figure 8 shows a part of the specification report for two anthropological cultural dimensions (1) Uncertainty avoidance and (2) Gender roles.

Every (pseudo natural language) sentence appearing in the report is clickable and is a hyperlink to an explanation page with more details. For example, the sentence “Website with LocalizationLevel [Broad]” (shown in Fig. 5) is a clickable hyperlink to a particular page explaining this role, shown in Fig. ‎9.

Fig. 9.
figure 9

Page providing more information

Full size image

5 Conclusion

In this paper, we proposed the Localized Website Design Advisor (LWDA) tool. LWDA dynamically generates advice on how to localize a website for a target country, language, level of localization, and website domain.

LWDA is based on two cultural information repositories, (1) The Localization Ontology and (2) the Cultural Markers Knowledge Base. We distinguished between them because each one has its own contribution and role. The ontology is used for providing abstract cultural specifications, while the knowledge base is used for storing the values and the description of these specifications for a particular website domain, a particular target country, and a specific website localization level. Distinguishing between the two makes the system more flexible and easier to maintain as it distinguish between information that may change over time (the knowledge base) and information that is more stable (the ontology) and only should be change when research proves that it is outdated.

LWDA was developed because there was a lack of a consolidated single source describing the cultural markers that need to be taken into consideration for the design of localized websites. Furthermore, LWDA’s locaization guidelines can also be used to evaluate existing websites.