Knowledge access, creation and transfer in e-government portals

The Authors

Dion Hoe-Lian Goh, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Alton Yeow-Kuan Chua, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Brendan Luyt, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Chei Sian Lee, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Acknowledgements

The authors wish to thank Audrey Charlotte Xavier, Syeda Rabia Fathima and Vineeta Jacob for their assistance in carrying out this research.

Abstract

Purpose – Knowledge management (KM) is an important consideration in e-government portals to ensure that knowledge flows efficiently between governments, individuals and organisations. A crucial aspect of e-government portals that has not been addressed adequately is the extent to which KM mechanisms have been implemented. Specifically, the authors argue that appropriate KM mechanisms are necessary to support the access, creation and transfer of knowledge between these portals and their users. The paper aims to propose an evaluation model for this purpose by first defining the main KM mechanisms and then burrowing deeper into their constituent dimensions.

Design/methodology/approach – An evaluation model known as knowledge access, creation and transfer (K-ACT) is presented which identifies three KM mechanisms for portals: knowledge access, creation and transfer. Each mechanism is characterised by a set of dimensions and sub-dimensions representing the tools and features for supporting that mechanism. The model was derived from an analysis of the literature and validated by two independent reviewers who were trained in information science, were familiar with the objectives of the project and understood the concepts underlying KM implementation in portals. Using this model, a checklist was developed and applied to 60 e-government portals in the Asian and North American regions to investigate the extent to which these KM mechanisms have been implemented.

Findings – The findings indicate that, on average, e-government portals featured only about 36 per cent of the KM mechanisms described in the model. Furthermore, no significant differences in the implementation of the KM mechanisms were found between the two regions' portals. The evaluation also offered potential areas for improvement based on the K-ACT model.

Originality/value – The present work has developed an evaluation model known as K-ACT which can be used to assess KM implementation gaps in e-government portals. This model can also be generalised to other types of portals. The evaluation also provides insights into the state of KM processes in the portals of the Asian and North American regions.

Article Type:

Research paper

Keyword(s):

Knowledge management; Knowledge creation; Knowledge transfer; Online operations; Government.

Journal:

Online Information Review

Volume:

Number:

Year:

2008

pp:

348-369

Copyright ©

Emerald Group Publishing Limited

ISSN:

1468-4527

Introduction

Knowledge is widely recognised as a strategic asset in improving organisational performance. As one of the major management strategies of recent times, knowledge management (KM) has attracted the attention of organisations which aim to enhance their business efficiency, performance and competitiveness. Wiig (1997) emphasised the need to manage knowledge from an organisational perspective. According to Wiig (1997), the objectives of KM are to make the enterprise behave intelligently to realise the best value of its knowledge assets, and to secure its viability and overall success. To meet these objectives, knowledge must be created, transformed and eventually disseminated and shared (Smith, 2001).

Essentially, governments are large organisations that must deal with significant amounts of information. Many governments have actively launched KM projects to meet the needs of the public, who are increasingly demanding high standards of quality, courtesy and responsiveness. In conjunction with such efforts are e-government initiatives that promote more efficient processes by facilitating improved access, via information technology, to information and services, and at the same time fostering better relationships with their citizens, businesses and other organisations. Here, KM plays a key role in ensuring that knowledge flows efficiently between the government and these groups.

Countries where e-government initiatives are implemented typically employ web portals as the gateway to the government and its services. Owing to the lack of social cues and interpersonal interaction on the web (Nah et al., 2005), the quality of knowledge and services delivered to the public and knowledge related to customer relationship management (CRM) becomes even more important.

There is much literature on e-government portals and KM, with topics such as the importance of KM in e-government (Harman and Brelade, 2001; Zahavora and Zelmene, 2004), KM strategies for the public sector (Misra et al., 2003), the technological infrastructure needed to support KM (Heck and Rogger, 2004), the structure of such portals (Everisto and Kim, 2005) and the effectiveness of service delivery through such portals (Daniel and Ward, 2006).

A crucial aspect that has not been addressed adequately is the extent to which KM mechanisms have been implemented in e-government portals. We propose an evaluation framework for this purpose by first defining the main KM mechanisms and then burrowing deeper into their constituent dimensions. Using this framework, we constructed a checklist and demonstrate its utility by evaluating 60 English-language official e-government portals from the North American and Asian regions. We also examine differences in the extent to which KM mechanisms in the portals were implemented in these two regions.

Related work

Knowledge management, customer relationship management and e-government

e-Government involves the use of internet-based technologies to provide convenient access to government information and services, and to open new ways of engaging citizens and enterprises. However, e-government initiatives are not without challenges. For one, as people become more accustomed to the convenience and speed at which these interactions take place, the demand for better and even faster access to information and integrated service delivery also rises. Furthermore, the exponential growth of transactions represents an opportunity for governments to study usage patterns, understand users and prepare for the future. KM is therefore seen as a compelling strategy to dovetail e-government initiatives to achieve at least three objectives. First is to cater to users' increasing expectations for better service delivery and information access (Harman and Brelade, 2001). Second is the move towards an anticipatory government that is forward-looking (Misra et al., 2003). Third is to facilitate interactions between all entities of the e-government and its constituents.

An important component in an e-government initiative is CRM. Peng and Chen (2005) define CRM as an information system that helps enterprises manage customer relationships and they assert CRM's importance in e-government as a mode of managing various governmental functions to increase customer satisfaction and loyalty. In addition, Accenture's (2003) study of e-government portals found that CRM underpins e-governments and that this belief was growing among government executives. The study emphasises that “as governments rethink their strategies to focus on delivering value, they must also create a customer impact” (Accenture, 2003, p. 4).

The CRM strategy thus goes hand in hand with the customer-centric approach of e-government. One benefit of applying CRM principles to e-government is that services provided to customers can be aligned with their preferences. Furthermore, CRM allows for better integration, transparency and access to information for the customer. This, in part, helps build closer ties between the government and its customers.

Zahavora and Zelmene (2004) note that there is a consensus among analysts, researchers and industry practitioners that customer relations, service and support are areas where the application of KM could bring significant benefits to an organisation since many CRM processes rely on knowledge resources. Gebert et al. (2002) classify knowledge flows in CRM processes into three categories:

Knowledge for customers. This is necessary to satisfy the knowledge needs of customers. An example is knowledge about products.
Knowledge about customers. This type of knowledge is accumulated to understand motivations of customers and to address them in a personalised way. This can include customer histories, expectations and so on.
Knowledge from customers. Here, customer knowledge about a product, service or other information is gathered and the organisation can use this knowledge to improve service or develop products.

CRM is intrinsic to viewing the government process as a business process and in treating all entities associated with the government as “customers”. KM can be utilised as part of CRM to improve customer-related processes such as identifying customer needs and wants, measuring satisfaction, selecting or improving channels for distribution of information, delivering services, ensuring quality of service and responding to enquiries (Zahavora and Zelmene, 2004). All these processes are relevant in e-government initiatives (Peng and Chen, 2005), hence these initiatives should focus on knowledge flows for, about and from government and their customers.

e-Government portals

A web portal is the typical implementation platform of an e-government initiative, providing access to the government and its services. Yong and Koon (2003) discuss how governments keen on adopting the right business models for e-government initiatives have made the “portal model” the predominant choice. This is because portals offer a single point of entry to multiple agencies and accord users the opportunity to interact easily and seamlessly with these agencies.

Much literature exists on the study of the nature and classification of portals. Dias (2001) classified portals by their functions as:

horizontal or public portals; and
vertical, corporate or enterprise portals.

Dias emphasises two main functions of a corporate portal: decision support and collaborative processing. These functions connect users to all available information and to everyone they need in order to accomplish their goals. Examples of portals belonging to this category are knowledge portals and enterprise information portals.

Daniel and Ward (2006) argue that e-government portals are similar to enterprise information portals since they provide a variety of features such as improved access to information, increased collaboration, greater use of existing applications and effective integration between applications. In addition, e-government portals allow users access to and interaction with a range of internal and external applications and information sources. An ideal e-government portal is often referred to as a “one-stop” portal (Tambouris and Wimmer, 2005) and is a direct consequence of the integrated delivery of information and services to customers of the government. Such portals assume the key attributes of enterprise information portals, enabling the government to unlock internally and externally stored information, providing users with a single gateway to personalised information necessary to make informed decisions and enabling collaboration (Dias, 2001).

e-Government portals, as one-stop portals, therefore will typically offer the following features: intuitive classification and searching, information subscription, information aids, collaborative information sharing, security and privacy, customisation and personalisation, information directories, multiple information delivery channels, access to a variety of information and services, connectivity to different departmental services via a single access point, and retrieval of services based on need and without knowledge of underlying departments (Dias, 2001; Tambouris and Wimmer, 2005).

KM models for the internet

Governments can increase the effectiveness of e-government initiatives via their portals by treating various entities as “customers”. KM becomes an important component in this by providing the mechanisms for creating, capturing, transforming and using knowledge. In addition, e-government initiatives must facilitate collaboration and knowledge flows between themselves and their users, and also between users. In this section, we review three recent KM models for the internet that attempt to capture these mechanisms in portals and other web sites.

As its name suggests, the acquisition, organisation and distribution (AOD) model (Schwartz et al., 2000) identifies three components for internet-based KM:

Acquisition. Collecting and storing knowledge from members of the organisation or other resources.
Organisation. Structuring, indexing and formatting the acquired knowledge.
Distribution. Retrieving relevant knowledge for the person who needs it at the right time.

The focus of this model is on how organisational knowledge can be delivered to the necessary point of action through multiple electronic delivery channels (push or pull mechanisms), e-mail and instant messaging. The AOD model however, lacks a cross-flow of knowledge across organisational boundaries, such as between the organisation and its customers.

Tiwana (2000) argues that successful web sites are centred on managing customer knowledge and proposes a customer knowledge cycle that comprises three parallel phases:

Acquisition. Development and creation of insights, skills and relationships supported by data capture tools and information technology.
Sharing. Disseminating and making available what is already known.
Utilisation. Integrating learning into the organisation by applying (at least in part) whatever is broadly available throughout the organisation to new situations.

Building on Tiwana's (2000) work, Nah et al. (2005) identify a need for both knowledge acquisition and dissemination to occur between organisations and their customers. They considered Davenport and Prusak's (1998) outline of KM processes which include knowledge generation, knowledge codification and knowledge transfer. They view the transfer of knowledge across organisational boundaries as desirable to facilitate the exchange and sharing of information between user communities and the organisation. The resulting internet-based customer KM model resembles Tiwana's (2000) model but replaces the knowledge utilisation component with knowledge dissemination because the former is internal to the organisation while the latter suggests both internal and external knowledge flows. The components of this model are:

Knowledge dissemination. Delivering knowledge to potential customers.
Knowledge acquisition. Acquisition of customer knowledge.
Knowledge sharing. Enabling mechanisms for users to provide online knowledge to peers.

Knowledge access, creation and transfer (K-ACT): a model for KM in web portals

Overview

We present a model that identifies the relevant dimensions of KM mechanisms in web portals as shown in Table I. This work is part of a larger effort aimed at investigating KM implementations on the web in various domains, including e-government.

We begin with the observation that while the internet-based customer KM model (Nah et al., 2005) has its strengths when compared to the other models (Schwartz et al., 2000; Tiwana, 2000), it is more applicable to e-commerce web sites and does not include many KM mechanisms that could be potentially useful to e-government and elsewhere. Our knowledge access, creation and transfer (K-ACT) model attempts to extend on existing work in this area by proposing three mechanisms for KM in web portals:

Knowledge access. It is the mechanism through which users obtain access to the knowledge in the portal. The emphasis is on the tools that support the individual user's acquisition of knowledge from the portal and it replaces the distribution, sharing and dissemination components of the other models that suggest a mix of portal-to-individual and individual-to-individual-via-portal approaches to sharing knowledge.
Knowledge creation. It expands on the acquisition component of Nah et al. (2005) to include both the acquisition of knowledge about the user and the acquisition of knowledge from the user. While the knowledge access mechanism provides tools for portal-to-user knowledge flows, this component supports user-to-portal flows.
Knowledge transfer. It supports user-to-user flows of knowledge. Here, the focus is on tools for knowledge sharing among individuals and organisations that have access to the portal. Like knowledge access, this mechanism distinguishes between the parties involved in knowledge sharing, unlike the components in the other models reviewed earlier.

Each mechanism in the K-ACT model is further described by a set of dimensions and sub-dimensions (where necessary) representing the tools and features for supporting that mechanism. An elaboration of these dimensions together with the rationale for their inclusion is presented in the following sections.

Knowledge access

Knowledge access provides support for searching, browsing, various forms of information delivery (Davies et al., 2005) and personalisation. The dimensions and sub-dimensions are:

Access to portal. This dimension assesses the ease of locating the portal using major search engines and directories. Visibility of a portal to potential users is essential to promote its successful use and a “good” portal in terms of this dimension is one that receives maximum exposure in major search engines and directories. In the present work, these refer to Google, MSN and Yahoo.
Search. This dimension evaluates the information retrieval capability of the portal. Search is a standard feature available on a majority of portals and other web sites allowing users to quickly find appropriate services and relevant information. This is sometimes preferred to browsing, especially in large portals. The search dimension is further divided into two sub-dimensions:
- Query. This refers to the types of querying services available, such as full-text search, advanced search and recommendations which assist users to locate information based on other likeminded users (Lau and Goh, 2006; Nielsen, 2000).
- Results display. This is the availability of options in the search results pages, such as sorting and search refinement.
Browse. In contrast to search, browsing facilitates content exploration given one or more organisational structures (Lau and Goh, 2006). Commonly used structures include sitemaps, keyword indices and hierarchical subject lists (Nielsen, 2000).
Personalisation. This dimension assesses the ability to personalise portal content and includes features commonly grouped under customisation (Jasco, 1999). Personalisation (Reneker and Buntzen, 2000) enables the creation of user accounts and profiles for targeted information delivery so that individuals can select and organise the information being presented. Users are also able to specify their preferred portal look-and-feel including fonts, colours and layout. Personalisation is further divided into two sub-dimensions:
- User-driven personalisation. This refers to the options available for personalising content based on individual needs.
- System-driven personalisation. This refers to the portal's capability to personalise content based on anticipated user types. For example, content can be packaged differently for individual citizens, potential immigrants and businesses.
Accessibility. The World Wide Web Consortium (W3C, 2005) defines web accessibility as support for people with disabilities to “perceive, understand, navigate, and interact with the web”. It is estimated that about 10 per cent of internet users have special needs (Seeman, 2004) and because governments have the responsibility to enable greater participation in society by such people, extending this tenet to the web is imperative. Examples of disabilities and special needs include an inability to see, hear, read or understand some or all of the information, and the inability to use the mouse (Seeman, 2004). Such problems can possibly be overcome by providing various interface modes (e.g. speech recognition), alternatives to audio-visual information and so on. The accessibility dimension assesses the extent of support for people with special needs to use the portal.
Information presentation. Information can be presented to the users in various forms such as text, graphics, audio and video. A judicious mix of different multimedia helps users obtain a better understanding of the portal's content (Grace et al., 2000). This dimension is concerned with the range of multimedia content available on the portal.

Knowledge creation

Knowledge creation involves a continuous and cumulative process of capturing customer information such as demographics, preferences and behavioural patterns for improving future business value (Bardzki and Reid, 2004). The relevant dimensions include:

User information acquisition. This dimension examines the features available to acquire information from and about portal users. Membership sign-up is perhaps the most common tool and provides a convenient way of capturing user behaviour within the portal (Nah et al., 2005). The type of information captured varies but may include demographics, contact information and preferences. Such information enables the organisation to construct customer profiles that support targeted communication (Rowley, 2004).
Feedback. This refers to the features used to acquire user feedback about the portal and its content. Examples include online forms and e-mail. Feedback helps an organisation assess its ability to meet the needs of its users and in the process, provide better value-added products and services (Rowley, 2004).
Domain data acquisition. In contrast to feedback which typically acquires general information, this dimension assesses the capability to elicit specific subject or domain information from users. Common examples include surveys and polls (Nah et al., 2002).

Knowledge transfer

Knowledge transfer or sharing involves the dissemination of individual experiences, information or knowledge to those who might need it (Williamson and Beghtol, 2003). Such sharing can also help the organisation understand its users better and facilitate the delivery of services in a more effective, efficient and responsive fashion. The dimensions and sub-dimensions are:

Online collaboration. This dimension evaluates the support provided by the portal for interaction and knowledge sharing (Wagner, 2005). The dimension is further divided into the following sub-dimensions:
- Organisation-to-user collaboration. This may either be synchronous (e.g. chat) or asynchronous (e.g. e-mail), but both involve representatives within the organisation who are able to assist portal users.
- User-to-user collaboration. Having a platform for users to share their ideas and opinions leads to a wealth of knowledge and enables an organisation to better understand its users. Examples include discussion forums, blogs and wikis.
Information alerts. This refers to the support for targeted information delivery to portal users. Alerts and reminders keep users informed about the latest information service or product they might be interested in Nah et al. (2005).
User support. This refers to support provided to users when problems arise during use. This is an integral feature in any portal because the level of user experience can vary greatly (Nielsen, 1994). Examples include FAQs, help desk support, online help and tutorials.
Resource sharing. Information found in the portal may be produced in-house, extracted from external sources (via links to other web sites) or contributed by users. Information from the latter two categories may be considered forms of knowledge sharing and this dimension examines the types of information that can be found within a portal.

Evaluating e-government portals using the K-ACT model

Developing the K-ACT checklist

Using the K-ACT model, we developed a checklist as an instrument for evaluating the extent to which KM mechanisms are implemented in portals. Items in the checklist corresponded to the model's dimensions and sub-dimensions. Checklists are considered valuable evaluation devices and have advantages such as reducing errors of omission, ensuring no feature overshadows another due to biases and being easy to understand and use (Goh et al., 2006; Stufflebeam, 2000).

The development of the checklist, including the elimination and inclusion of items, was guided by the following requirements and is consistent with our earlier preliminary work in this area (Lee et al., 2007). It had to:

be applicable across various domains (such as e-government, healthcare, etc.);
have comprehensive coverage of all areas identified in the K-ACT model; and
focus on functional and technical features that facilitate KM.

Given this scope, features such as content, usability and domain-specific services (e.g. integrated access to government services and transactions) were not included in the checklist. The checklist is found in Appendix 1.

Applying the K-ACT model to e-government portals

We used the K-ACT checklist to evaluate 60 English language e-government portals to determine to extent to which KM mechanisms were implemented. The portals were from the Asian and North American regions as we also aimed to identify differences in KM implementations in these two regions. The list of portals evaluated is found in Appendix 2.

The United Nation's (2005) Global E-Government Readiness Report was used to select the portals. In the report, only five Asian countries featured in the top 25 countries. USA topped the list, while Canada followed in the fifth position. We therefore expected North American portals to perform better in their implementation of KM mechanisms than Asian portals. Only official government portals were selected. We randomly selected one official portal each from 30 countries in the Asian region and one official portal each from 30 randomly selected states/provinces from the USA and Canada in the North American region, plus their respective federal government-level portals.

Each of the 60 portals was evaluated by two reviewers independently. The reviewers were trained in information science, were familiar with the objectives of the project and understood the concepts underlying KM implementation in portals. Every item in each dimension/sub-dimension was marked “Yes” if it was found in the portal or “No” if not found. The items present were summed to obtain the level of support for that dimension/sub-dimension. The scores for each portal were then compared. Any discrepancies in the evaluation were addressed by a combined re-assessment between the two reviewers. Instances of discrepancies were small as reflected by Cohen's (1960) Kappa measure, which was calculated to be 0.954, indicating very high levels of agreement between the reviewers.

Findings and analyses

This section presents the results of our evaluation of the 60 e-government portals. Owing to space constraints, analyses of individual portals are not possible and the discussion will focus on comparisons between North American and Asian portals.

Knowledge access

Table II shows that the maximum possible score for the knowledge access mechanism was 22. The mean score for North American portals was 7.70, indicating that on average portals only implemented approximately eight of the 22 features, or a difference of 65 per cent from the maximum score. Put differently, only 35 per cent of knowledge access features were found to be available in North American portals. The best performing North American portal was the Government of Ontario (18). The mean score for Asian portals was a little higher at 8.13, suggesting that collectively about only 37 per cent of knowledge access features were implemented. Here, the Singapore government portal was the best performer (16). However, differences between the two regions in the level of support for knowledge access were statistically non-significant (t(58)=0.509, p=0.613).

A more detailed analysis of the dimensions within knowledge access (Table III) yielded the following observations:

Access to portal. The North American portals had perfect scores while in the Asian region only the North Korean government portal did not fulfil this criterion.
Query. Eight portals in the Asian region obtained a score of zero in this sub-dimension. The North American region fared better with only one portal scoring zero.
Results display. More Asian portals (21) scored a zero in this sub-dimension compared to North American portals (nine).
Browse. Here, the Asian portals fared better with an average score of 1.43 out of a maximum of three features as opposed to 1.03 for North American portals. Only one Asian portal scored a zero compared to eight North American portals.
User-driven personalisation. Portals in neither region performed well, with 22 North American portals and 18 Asian portals obtaining zero, indicating a clear lack of personalisation options for users.
System-driven personalisation. Scores were similarly low, with 19 portals in each region scoring zero.
Accessibility. Disappointingly, most portals only provided minimal support for the hearing and visually impaired. None obtained a maximum score, while a surprising ten portals from North America scored zero compared to only four from Asia.
Information presentation. Again, most portals had limited ways of presenting information to users. Of multimedia types, images were the most popular, followed by animation. Only four Asian portals and one North American portal were found to support all multimedia types.

Knowledge creation

As shown in Table II, the maximum score for the knowledge creation mechanism was six. The mean score for North American portals was 1.50, suggesting that only 25 per cent of the features in this mechanism were implemented. The portal of the State of Maine performed best in this region, with a score of four features implemented. Portals in Asia fared better with a mean score of 2.37, indicating that about 40 per cent of all features in the knowledge creation mechanism were implemented. Here, the portals from the governments of Dubai, India and Malaysia shared the highest score of five. The difference between the two regions in the level of support for knowledge creation was statistically significant (t(58)=2.72, p=0.009) and it appears that governments in the Asian region are attempting to understand their users and acquire knowledge from and/or about them more than governments in the North American region.

Analysis of the dimension and sub-dimension scores (Table III) revealed the following:

User information acquisition. Asian portals performed better, with 13 portals obtaining maximum scores compared to only four North American portals. Furthermore, 16 North American portals obtained a score of zero as opposed to only ten Asian portals.
Feedback. The Asian and North American regions had comparable numbers of portals with maximum scores in this dimension (26 versus 24), suggesting that most e-government portals understand the importance of user feedback.
Domain data acquisition. Portals from neither region obtained the maximum (3.33) score here. The Asian region performed better with five portals obtaining a score of 2.22. About 27 North American portals obtained a score of zero compared to 23 Asian portals.

Knowledge transfer

Portals from both regions performed poorly in knowledge transfer (Table II). While the maximum score possible was 22, the mean score for North American portals was 5.56, indicating that only about 25 per cent of knowledge transfer features were available. The portal of the US Federal Government (FirstGov.gov) performed best with a score of 10. On the other hand, portals in Asia performed worse with a mean score of 4.9. This means that only approximately 22 per cent of knowledge transfer features were found during the evaluation. Here, the Australian government portal scored highest at 11. However, on the whole, differences between the two regions were statistically non-significant (t(58)=−1.07, p=0.288), suggesting similar levels of support for knowledge transfer in the two regions' portals.

Further analysis of the dimension and sub-dimension scores (Table III) revealed the following:

Organisation-to-user collaboration. This was poorly supported in both regions with only five North American and three Asian portals scoring maximum points and the remainder scoring zero.
User-to-user collaboration. Features in this sub-dimension were once again found to be very poorly supported. All 30 North American portals scored zero here. Interestingly, there were four Asian portals that had one feature implemented in this sub-dimension. The portals of the Bhutan and Philippines governments had discussion forums, the Cambodian government portal supported blogs, while the Malaysian government portal had instant messaging.
Information alerts. Only one Asian portal (Singapore) obtained the maximum score while none in the North American region did. Seven North American and Asian portals did not provide features supporting this dimension.
User support. No portals obtained the maximum score in this dimension but in general North American portals had better user support than their Asian counterparts. For example, eight North American portals had the next-highest score of four as opposed to only one Asian portal. In addition, five North American portals scored a zero compared to ten Asian portals.
Resource sharing. Three Asian portals and one North American portal obtained the maximum score while there were no portals scoring zero, indicating at least basic levels of support for sharing information among users.

Overall implementation support for the K-ACT model

Taking all KM mechanism scores into consideration, North American portals were found to implement only 29.52 per cent of all KM features, while Asian portals were only slightly better at 30.80 per cent. The differences in values were statistically non-significant (t(58)=0.437, p=0.664) indicating that e-government portals from both regions were at the same stage of KM implementation.

Table IV shows a summary of the e-government portals organised by the percentage of KM mechanisms implemented. Most portals (39) implemented between 26 and 50 per cent of all features described by the K-ACT model. A minority (three) implemented up to 75 per cent, while no e-government portal implemented more than 75 per cent of all features. A χ ² analysis revealed no significant differences in implementation levels between Asian and North American portals (χ ²(2, N=60)=0.786, p=0.68), again demonstrating that even with this deeper analysis, portals in both regions show similar levels of support for the K-ACT model.

The top e-government portals were, in descending order of percentage implementation scores, Singapore (62 per cent), the Province of Ontario (56 per cent), Malaysia (58.50 per cent), and Australia, the US Federal Government and the Ministry of Foreign Affairs in Japan (all 48 per cent). See Appendix 2 for the complete list of scores. Singapore's e-government portal scored well in the knowledge access and knowledge creation mechanisms, having on average at least half the number of checklist items found in each dimension and sub-dimension. However, it performed poorly in the knowledge transfer mechanism, scoring zero in the organisation-to-user and user-to-user collaboration sub-dimensions. There was also minimal user support, suggesting that users had little help when they encountered problems while using the portal. The portal obtained the maximum score in the information alerts sub-dimension, indicating that targeted information delivery was the main mode of knowledge transfer.

A lack of knowledge transfer support was the main weakness among the other five portals mentioned above as well. However, there were some notable differences – the Australian e-government portal provided an ask-an-expert service, while the Malaysian e-government portal supported instant messaging. In addition the two North American portals provided better user support than the Asian portals in terms of the number of checklist items found in that dimension.

The bottom e-government portals were, in descending order of implementation scores, Israel and Brunei (both 16 per cent), North Korea and the Government of Nunavut (both 14 per cent), and Yemen and the Province of New Brunswick (both 12 per cent). These portals were characterised by very low scores across all dimensions and sub-dimensions of the K-ACT model and in particular had zero scores in results display, user-driven personalisation, system-driven personalisation, domain data acquisition, organisation-to-user collaboration and user-to-user collaboration. The e-government portal of Yemen and the Province of New Brunswick, the worst performers, also had zero scores in querying, feedback and user support.

Apart from the low scores in all aspects of the K-ACT model, some interesting observations of these portals include:

only the e-government portals of the provinces of New Brunswick and Nunavut (Canada) supported full-text searching;
the Nunavut portal did not support browsing; and
multilingual support was available in the bottom five portals.

Discussion and conclusion

KM is a crucial consideration in e-government portals to ensure that knowledge flows efficiently between governments, citizens and organisations. In this paper, we contribute to the body of literature by developing an evaluation framework known as K-ACT which identifies three KM mechanisms for portals: knowledge access, creation and transfer. Each mechanism is further characterised by a set of dimensions and sub-dimensions representing the tools and features for supporting that mechanism. Owing to the diversity inherent in the field of KM, the K-ACT model attempts to distil important concepts that are relevant to evaluating KM implementation levels in portals. In our work, this is achieved through a thorough review and analysis of related literature. Using this model, we constructed a checklist and demonstrate how it can be applied by evaluating 60 e-government portals in the Asian and North American regions. The evaluation provided insights into the state of KM processes in the portals of the two regions and also showed potential areas of improvement with reference to the K-ACT model.

In general, our evaluation found no significant differences between portals from either region in their implementation of KM. Despite the expectation that North American portals would outperform Asian portals due to the state of North American technological infrastructure development, the level of KM implementation in all portals, as described by the K-ACT model, were generally found to be low. On average portals implemented less than 40 per cent of all features described in each mechanism. This suggests that KM is still a nascent concept in e-government. Nevertheless, it is noteworthy that, among all mechanisms, a statistically significant difference in KM implementation levels was found in knowledge creation, with Asian portals leading the effort to generate knowledge. Given this, two possible follow-up studies could be conducted:

whether these differences are due to a better awareness of KM processes in the Asian region or due to more stringent privacy policies in the North American region, leading to a limit in the types of information that could be captured from users; and
whether and how such knowledge is used to improve the service levels of these portals.

Despite these results it is important to note the progress made by governments to provide more services online via their portals (United Nations, 2005). However, we argue that, parallel to such work, KM processes should also be considered to ensure that these portals are user-centric. As discussed, KM can be used as a part a of CRM strategy to identify needs and wants, measure satisfaction and quality of service, design services and respond to enquiries. The K-ACT model can therefore be used as a guide to determine KM implementation gaps in e-government portals. Stakeholders should, however, decide whether these gaps should be addressed, given the objectives and scope of their respective portals. In addition, because the scope of the K-ACT model is deliberately focused on features that promote KM practices in portals, it does not attempt to address the universe of KM or software engineering-related concerns or problems. However, design and usability issues are sufficiently addressed in the literature (Barnes and Vidgen, 2003), so developers should use the K-ACT model in conjunction with established usability instruments and guidelines during portal implementation.

The present research has some limitations that could be addressed in future work. First, only English language portals in the North American and Asian regions were studied. This set of portals is not representative of the entire population of portals around the globe. Non-English portals were excluded due to the language barrier. A more comprehensive study can therefore be undertaken to evaluate portals from other regions including South America, Europe and Africa. Doing so will paint a better picture of the extent to which KM mechanisms have been employed in e-government portals. Second, while the current K-ACT model has utility as an evaluation tool, the work here represents only a first step in measuring KM implementation levels in portals. Future work will need to examine the validity and reliability of the model, with the objective of developing a robust instrument for portal evaluation.

As this is a preliminary study, future models could be built to incorporate other elements of e-government such as electronic voting, social security management, registration of births and deaths and so on. Underlying all of these functions is the issue of privacy, an important consideration for governments worldwide today. How well e-government portals protect individual privacy and how well they inform the public about what privacy rights they have should be taken into consideration in a more elaborate model for portal evaluation.

Finally, it must be admitted that, for some individuals, accessing government information online will be problematic. There has been much discussion of the digital divide and the fact that, for certain groups of people, internet access is not an easy matter. For example, the chances of possessing the resources or skills needed to access the internet are not as good for older or less educated people, or those from a lower socio-economic class. For these people, provision needs to be made for off-line access to government services as well. The existence and clear presentation of such mundane things as phone numbers and regular mail addresses should also be part of an updated K-ACT model.

Table IThe K-ACT model

Table IIKM mechanism scores

Table IIIDimension/ sub-dimension scores

Table IVPercentage of KM mechanisms implemented

Table AI

References

Accenture (2003), eGovernment Leadership: Engaging the Customer, available at: www.accenture.com/xdoc/en/newsroom/epresskit/egovernment/egov_epress.pdf, .

Knowledge access, creation and transfer in e-government portals

The Authors

Dion Hoe-Lian Goh, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Alton Yeow-Kuan Chua, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Brendan Luyt, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Chei Sian Lee, Division of Information Studies, Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore

Acknowledgements

Abstract

Article Type:

Keyword(s):

Journal:

Volume:

Number:

Year:

pp:

Copyright ©

ISSN:

Introduction

Related work

Knowledge management, customer relationship management and e-government

e-Government portals

KM models for the internet

Knowledge access, creation and transfer (K-ACT): a model for KM in web portals

Overview

Knowledge access

Knowledge creation

Knowledge transfer

Evaluating e-government portals using the K-ACT model

Developing the K-ACT checklist

Applying the K-ACT model to e-government portals

Findings and analyses

Knowledge access

Knowledge creation

Knowledge transfer

Overall implementation support for the K-ACT model

Discussion and conclusion

References

Appendix 1. The K-ACT checklist

Appendix 2. Portals evaluated and their implementation scores

Corresponding author