upa - home page JUS - Journal of usability studies
An international peer-reviewed journal

Usability Evaluation of a Tag-Based Interface

Rajinesh Ravendran, Ian MacColl, and Michael Docherty

Journal of Usability Studies, Volume 7, Issue 4, August 2012, pp. 143 - 160

Article Contents


Tags are a Web 2.0 technology, enabling users to assign keywords to Web resources (e.g., photo, video, people, etc.) primarily for the purpose of information management (see Figure 1; Flickr is a website that allows users to manage and share photographs online). Tags are a popular and easy-to-use technology (Smith, 2007); personal, contextual, and dynamic (Marlow, Naaman, Boyd, & Davis, 2006); and a source of knowledge about users (Durao & Dolog, 2009). Tags aid users to recall and retrieve information content, and when represented as tag clouds (see Figure 1 and 2), they facilitate visual information retrieval (Hassan-Montero & Herrero-Solana, 2006). Tags can also help to establish associations between like minded individuals through analysis of tag semantics (Durao & Dolog, 2009; Qi Xin, Uddin, & Geun-Sik, 2010).

Figure 1

Figure 1. Flickr photo tag cloud (source: http://www.flickr.com/explore/, Reproduced with permission of Yahoo! Inc. (c)2012 Yahoo! Inc. FLICKR is a registered trademark of Yahoo! Inc.)

The tag-cloud shown in Figure 1 is weighted based on tag occurrence frequency. Frequency is represented through varying font sizes where a higher frequency has a larger font size and vice-versa.

Tags are widely used in the financial services domain to aid personal financial management (see Figure 2, a sample financial tag cloud) via either external tools such as Mint1 or bank provided tools such as Australia-New Zealand (ANZ) Bank’s MoneyManager2 service. These tools enable users to assign tags to annotate transactional data for purposes such as budgeting, expense tracking, cash flow analysis, etc. However, the tools only allow users to assign tags to financial transactions at a high level such as category or description, but not at a fine-grained level for resources such as bank accounts or billers. There may be compelling advantages to assigning tags at a lower level of detail, particularly in the online and mobile banking contexts, which could open the doors to tag-based interaction alongside personal financial management.

Figure 2. Financial tag cloud

The tag-cloud shown in Figure 2 is weighted based on tag transaction amount. Amount is represented through varying font sizes where a higher frequency has a larger font size and vice-versa.

The earlier mentioned characteristics of tags, particularly personal and contextual, appear suitable to afford customized interactions that are unique to every individual based on tags assigned to resources. We refer to this notion as tag-based interaction. To date, tags have yet to be studied in the online and mobile banking contexts to afford customized interactions. Such research is important as customization is cited as a key determinant of user satisfaction of online banking in Australia (Rahim & JieYing, 2009).

In this study, we explored the use of tags to facilitate customization based on a conceptual interaction customization framework put forward by Fung (2008).

We hypothesized that tag-based interaction can improve usability, especially the user satisfaction of online and mobile banking. Aside from usability, other motivations for this study include mobile banking adoption issues such as privacy and security (Wessels & Drennan, 2010). Through the use of personal tags, users may feel more comfortable carrying out banking transactions using their mobile devices especially in public places given that only tags are displayed on screen, which is less likely to be meaningful to other individuals.

The rest of this article is structured as follows. In the next section we provide background on our work. Then we outline the method and process of our evaluation. Next we report on the evaluation results and discuss the findings. Subsequently we present a list of challenges related to the tag-based interface/interaction and recommend a set of potential solutions. Finally we conclude the study and describe future work.


In this section, we introduce the range of taggable resources identified in online and mobile banking and the different types of customization supported via user-defined tags.

Taggable Resources

There are five types of key resources in online and mobile banking: bank account, biller, reference, message, and application (Ravendran, MacColl, & Docherty, 2011a, 2011b). These resources were identified through examination of the online and mobile banking websites of two leading banks3 in Australia. The examination specifically focused on personal banking as it appeals to a larger audience.

Table 1. Taggable Resources

Table 1


We adopted the conceptual interaction customization model based on human-to-human interaction proposed by Fung (2008). In our earlier work (Ravendran et al., 2011b), we framed interaction customization as part of a broader work on customization as seen in the literature. This was important to both understand and distinguish interaction customization from other types of customization. Interaction customization encompasses three types: remembrance-based, comprehension-based, and association-based.

The following sections provide a brief definition of each interaction type and discuss the application of tags to facilitate these interactions in the banking context.


The remembrance-based customization is defined as customization through simple remembering of user’s information based on the recurrence rate of a particular action on a website (Fung, 2008). This customization can be fulfilled through tags assigned to resources that are presented as tag clouds (Ravendran et al., 2011a, 2011b). This provides a visual retrieval interface that can simplify and ease the execution of past or recurring transactions. Simply by clicking on a tag, related information about a transaction that the tag is associated with can be retrieved and displayed. If a selected tag is associated with two or more tags then the tag cloud can be filtered to show tags that are co-occurring with the selected tag. This removes the need to navigate to a different page or perform a manual search query. This also means, to carry out a past or recurring transaction, users will only need to update necessary information such as amount (if different) and possibly retain other details. The following scenario and Figure 3 is an example of remembrance-based customization in the online context.

Scenario: Recurring monthly mobile bill payment. User selects “mobile” (1) tag from tag cloud. As a result, the form is auto-completed and relevant tags are pre-selected (checkmark).

Figure 3

Figure 3. Recurring bill payment


The comprehension-based customization is defined as customization through recognition of user’s behaviors used to provide assistance towards fulfilling the user’s needs (Fung, 2008). This customization can be fulfilled by inferring possible banking actions (i.e., fund transfer) based on tags selected by a user (Ravendran et al., 2011a, 2011b). Such an inference is possible for tags with certain types of relations (e.g., account to account, account to biller). Using these relations and simple pre-defined rules (e.g., transfer from Savings account to Visa account is valid but not the other way around) possible actions can be populated. The actions may also constitute different types of services or features offered by the bank for a particular need that otherwise may not be known to the user. By providing users with relevant options, banks may be able to provide subtle suggestions to users based on their personal banking usage. For example, if a user performs fund transfers to a selected payee two months in a row, then a new option suggesting the user to schedule a monthly transfer may be provided. The following is an example of comprehension-based customization in the online context.

Scenario: New fund transfer from Everyday account to John’s account. User selects “Everyday” (1) and then “John” (2) tags. The possible actions are populated as “Transfer from Everyday to John” and “View transaction history of Everyday involving John”.

Figure 4

Figure 4. New fund transfer


The association-based customization is defined as customization through association of a user’s behaviors with other individuals who share similar interests or needs (Fung, 2008). This customization can be fulfilled by recommending tags to users (Ravendran et al., 2011a, 2011b). Tags can be associated and recommended to users based on certain criteria such as biller name or type. For example, when users select the biller Vodafone, a mobile service provider, tags associated with this specific biller can be recommended (i.e., phone, mobile, etc.). Tags associated with a particular resource can be either defined by users (folksonomy) or system (taxonomy), or both combined (automanual folksonomies; Smith, 2007). The relevance and appeal of recommended tags may be further improved by making sense out of the underlying meanings of tags via semantic analysis (Durao & Dolog, 2009; Qi Xin et al., 2010). This in turn could assist in forming more relevant associations between like-minded individuals within a community of users through discovery of semantic relationship between tags. As a result, recommendations that potentially deliver greater value are plausible, for example, a set of related services based on user’s banking usage. Below is an example of association-based customization in the online context.

Scenario: Tag recommendation for multiple bill payment (mobile and money transfer): User selects “Vodafone” (1) and then “OzForex” (2) biller tags, and then enters or selects a description (3). As a result, a set of related tags are shown that are used in the context of the selected billers.

Figure 5

Figure 5. Tag recommendation

1Mint, http://www.mint.com

2ANZ MoneyManager, http://www.anz.com/ANZ-moneymanager/default.asp

3Commonwealth Bank (http://commbank.com.au) and Suncorp Bank (http://www.suncorp.com.au)


Previous | Next