Int. J. Intelligent Information and Database Systems, Vol. 3, No. 1, 2009
Copyright © 2009 Inderscience Enterprises Ltd.
Personal knowledge management for knowledge
workers using social semantic technologies
Haklae Kim*, John G. Breslin and
Digital Enterprise Research Institute,
National University Ireland, Galway,
IDA Business Park, Upper Newcastle, Galway, Ireland
Department of Business Administration,
AnSeo-dong, Chonan, Chungnam, Korea
Biomedical Knowledge Engineering Lab,
Seoul National University,
Jong-Ro Gu, Yeon-Gun Dong, Seoul, Korea
Abstract: Knowledge workers have different applications and resources in
heterogeneous environments for doing their knowledge tasks and they often
need to solve a problem through combining several resources. Typical personal
knowledge management (PKM) systems do not provide effective ways for
representing knowledge worker’s unstructured knowledge or idea. In order to
provide better knowledge activity for them, we implement Wiki-based sociAl
Network Thin client (WANT) that is a wiki-based semantic tagging system for
collaborative and communicative knowledge creation and maintenance for a
knowledge worker. And also, we suggest the social semantic cloud of tags
(SCOT) ontology to represent tag data at a semantic level and combine this
ontology in WANT. WANT supports a wide scope of social activities through
online mash-up services and interlink resources with desktop and web
environments. Our approach provides basic functionalities such as creating,
organising and searching knowledge at individual level, as well as enhances
social connections among knowledge workers based on their activities.
Keywords: semantic web; semantic wiki; personal knowledge management;
PKM; folksonomy; social tagging.
Personal knowledge management for knowledge workers 29
Reference to this paper should be made as follows: Kim, H., Breslin, J.G.,
Decker, S., Choi, J. and Kim, H. (2009) ‘Personal knowledge management for
knowledge workers using social semantic technologies’, Int. J. Intelligent
Information and Database Systems, Vol. 3, No. 1, pp.28–43.
Biographical notes: Haklae Kim is a Researcher at the Digital Enterprise
Research Institute at the National University of Ireland, Galway. His research
interests include the semantic web, folksonomies and social network.
John G. Breslin is a Researcher and an Adjunct Lecturer at the Digital
Enterprise Research Institute at the National University of Ireland, Galway. His
research interests include social software, online communities and the semantic
Stefan Decker is a Professor and Director of the Digital Enterprise Research
Institute at the National University of Ireland, Galway. His research interests
include the semantic web, digital libraries and social semantic desktop.
Jaehwa Choi is a Professor at Dankook University. His research interests
include knowledge management, social software, e-learning and business
Honggee Kim is a Professor at and Director of the Biomedical Knowledge
Engineering Lab at Seoul National University. His research interests include
the semantic web, medical knowledge engineering, and knowledge
The goal of personal knowledge management (PKM) aims to make knowledge workers
better at capturing, sharing and using knowledge and maximising their personal
effectiveness in the social aspect of their jobs (KM Magazine, 2004). Knowledge workers
has been using desktop and web applications to capture ideas or thoughts and to manage
schedules, addresses or tasks, etc. There are many different applications such as e-mail
clients, word processors and web browsers for knowledge worker’s daily routine. But
since knowledge is scattered across applications and websites, it is not easy to aggregate
or combine a right set of knowledge and operations required for their specific tasks for
knowledge workers. In addition, knowledge can be evolved by social interaction among
knowledge workers; knowledge worker’s activities have social characteristics such as
connecting, communicating, and collaborating with others. In this sense, PKM should
support both individual and social level for knowledge activities. Traditional PKM
systems tend to be suitable for supporting specific functions such as managing schedule
and address for individuals. This makes it difficult to extend and combine to other
applications or services.
There are emerging trends associated with computing environments that support
personal knowledge activities. ‘Web 2.0’ comprises of technologies and services
to enable users to collaborate and share social contents. They include social
software, content syndication, messaging protocols such as weblogs, wikis,
podcasts, really simple syndication (RSS) feeds, etc. The majority of popular Web 2.0
H. Kim et al.
sites, Flickr (http://www.flickr.com), del.icio.us (http://del.icio.us) and Technorati
(http://www.technorati.com), are connecting people into communities creating networks
of shared experiences using folksonomy and RSS. ‘Social semantic desktop’ can provide
reliable technologies to enhance functionalities of PKM. The social semantic desktop is a
new computing paradigm that provides an advanced way to create, automate and
structure information and the technology convergence including the social network,
community services and P2P services (Decker and Frank, 2004). It could provide a
transformation of a typical desktop system into a collaborative environment that supports
both personal computing and information sharing via social channels.
PKM is not only focused on managing data, but also on connecting people and
sharing data among them. New social and semantic technologies can be able to provide
knowledge workers to organise their thoughts and ideas in a relevant, timely manner. In
this paper, we focus on social tagging as a way of representation and sharing for
knowledge worker’s ideas or thoughts and wiki for organising their daily tasks. Social
tagging can be an effective method to support, extend or derive values from human social
behaviour. Wiki allows knowledge workers to make their internal knowledge more
explicit and more formal. A proposed system aims to combine both social tagging and
wiki features to improve effectiveness and efficiency for various activities of knowledge
This paper is organised as follows: Section 2 describes a type of knowledge and
collaborative tagging for social features of knowledge and summarise about semantic
wikis. Section 3 describes limitations of current tagging systems and introduces Social
Semantic Cloud of Tags (SCOT) – representation of tag data at a semantic level. Section
4 presents WANT – a wiki-based PKM system with design principles and its architecture.
Section 5 describes the main features of WANT and explains how tags can be mapped to
the SCOT ontology and we conclude in Section 6.
2 Related work
2.1 Types of knowledge
There are many definitions of knowledge and classifications or categorisations of
knowledge. According to Spender (1994), knowledge can be classified into two
dimensions: ‘explicit/tacit’ and ‘individual/social’ knowledge. Individual explicit
knowledge (conscious knowledge) is located in an individual in the form of facts,
documents and files that can be stored and represented from personal records. Individual
tacit knowledge (automatic knowledge) means tacit knowing, including practical
knowledge of people and performance of different types of skills. Social explicit
knowledge (objectified knowledge) represents the shared corpus of knowledge by
communities and social tacit knowledge is fundamentally embedded in the forms of
social practice (Nahapiet and Ghoshal, 1998).
A knowledge process, in general, transforms individual knowledge to social
knowledge. Knowledge is created through a social interaction in organisations or
communities where knowledge workers are involved. But, traditional knowledge
representation approaches constructed by domain experts (e.g., taxonomy, ontology)
provide strict structures with a high-level formality for describing knowledge. These
approaches are limited to represent individual knowledge worker’s unstructured thinking
and to support continuous feedbacks or interactions with others. We could come up with
a collaborative tagging (also known as folksonomy, social classification, social indexing)
as an alternative for knowledge representation and sharing from an individual knowledge
worker’s point of view.
Personal knowledge management for knowledge workers 31
2.2 Tagging for representing knowledge
Tagging is a way of representing concepts by cognitive association techniques, without
enforcing categorisation. A tagging system has been adopted in many social software
applications such as weblogs, social bookmarking and social networking sites. This
approach brings an important advantage to the knowledge workers in the form of a
simple way to describe their knowledge in individual information spaces and to share it in
online communities. A tag-a labelled keyword, is a type of metadata for a resource such
as a resource link, a web page, a picture, a blog post, etc. The resources can be tagged
with as many tags as desired because there are no restrictions on which or how many to
The result of participating tagging activities can be represented by folksonomies.
Folksonomies, a term first coined by Tomas Vander Wal in 2004, are user-generated and
distributed classification systems, emerging through bottom-up consensus (Merholz,
2004). The essence of folksonomies is user participation and internet-mediated social
interaction. The tags in folksonomies are chosen by knowledge workers and may be
reused and shared by other knowledge workers. Since a large number of users participate
in creating, adding and sharing metadata in the form of keywords, folksonomic tagging is
regarded as a social and democratic process (Golder and Huberman, 2006) and as a
collective and social knowledge. Quintarelli (2005) points out that, ‘without social
distributed environment that suggests aggregation, tags are just flat keywords’.
Knowledge workers do not necessarily have to be an expert, but can also be a creator
or consumer of the content. They can collaboratively create and manage tags to annotate
and categorise content. This activity establishes social connections among them and
improves social reinforcement.
2.3 Semantics in wiki
A number of subsequent attempts have been made to solve the limitations of knowledge
activities by various approaches. In particular, semantic web researchers have become
increasingly interested in studying wiki (Hepp et al., 2007). Although wiki systems
administrate collaborative contents, they only provide a limited number of functions for
structuring the contents. Content in typical wikis is encoded in HTML, making it difficult
to represent semantics for the content. A semantic wiki is a wiki system that has
an underlying semantic model of the knowledge described in its pages
(http://en.wikipedia.org/wiki/Semantic_wiki). These approaches aim to combine
semantic data into HTML contents and to enhance machine-readable performance. There
are several semantic wiki implementations such as Platypus Wiki (Campanini et al.,
2004), Rhizome (Souzis, 2005), Semantic MediaWiki (Volkel et al., 2006), etc.
H. Kim et al.
3 Social semantic tagging
Although a number of studies have been made on Web 2.0, little attention has been given
to Web 2.0 from a semantic web perspective. There is a gap between semantic web
research topics and Web 2.0 applications, since much semantic web research has thus far
been focused on developing standards and recommendations. On the other hand, Web 2.0
plays an important role by leading users to participate in online communities.
Technologies for Web 2.0, however, are not mature enough to deal with effective and
efficient services, in particular, those associated with the social tagging and folksonomies
(Gruber, 2007). For instance, a critical problem in typical tagging systems is that they do
not provide a uniform way to share and reuse tag data amongst users or communities.
Although most popular Web 2.0 sites such as del.icio.us and Flickr provide XML or
JSON-based data using open APIs, there is no uniform structure or semantics to represent
tag data. Therefore, it is not easy to meaningfully search, compare or merge ‘similar
collective tagging data’ (Tagcommons, 2007) from different sources. This makes it
difficult to share, reuse and integrate tag data among users or across different services.
From a knowledge worker’s point of view, the limitations can be a barrier to adopt
tag-based knowledge representation.
3.1 Overview of SCOT ontology
The SCOT ontology (http://scott-project.org) is an ontology for sharing and reusing tag
data and for representing social relations across different sources (Kim et al., 2008). It
provides the structure and semantics for describing resources, tags, and users and
provides extended tag information such as synonyms, spelling variants, tag frequencies,
tag cooccurrence frequencies and tag equivalence in order to reduce tag ambiguity. Our
approach follows the principle ‘a little semantic goes a long way’ (Hendler, 2007). The
ontology model is designed both with minimal structure and minimal semantics in a
simple RDF format. In order to share and reuse the data with other applications, the
ontology model provides a consistent method for sharing existing sets of tags amongst
3.2 The SCOT ontology model
The SCOT ontology generically models tagging activities for typical online communities
and relations between components (i.e., users, tags, resources, etc.) of the activity. We
recapitulate the formal model for a folksonomy introduced in Hotho et al. (2006). A
formal model of SCOT (S) is a tuple:
: ( , , , )U T R Y
U: set of users who participate in the tagging activity
T: set of tags that is assigned to resources
R: set of resources each of which has an indefinitely unchanged link that is called
Personal knowledge management for knowledge workers 33
Y: a ternary relation between U, T and R (i.e., Y ⊆ U × T × R), that represents
There is an implication that T has unique URIs for representing each tag for all resources.
For instance, del.icio.us and Flickr have their unique tag URIs such as
‘http://del.icio.us/tag’ and ‘http://www.flickr.com/photos/tags’, respectively. Using the
URI of T, we can connect and navigate the resources even if we do not have actual
resource information. From this perspective, a tag can be distinguished resources with a
keyword or a single term without a specific URI. R has an indefinitely unchanged link
that is called a ‘permalink’. Therefore, an individual tag and its URI in T are connected
with the permalinks of R on a web page.
Figure 1 shows a simplified model of the SCOT ontology with its top-level concepts
and with relations to other existing vocabularies. The concepts user, tag and resource for
the SCOT have links to FOAF (Brickley and Miller, 2005), SKOS (Brickley and Miles,
2005) and SIOC (Breslin et al., 2006), respectively.
Figure 1 SCOT ontology model (see online version for colours)
We use SIOC concepts to describe site information and relationships among
‘container-item’, ‘site-site’ and use FOAF concepts to represent a human or machine
agent as a tag. These can be generated either manually by a human user or automatically
by a machine. Also, the model attempts to represent the relationships among users. This
H. Kim et al.
relationship has two aspects: ‘agent-agent’ and ‘agent-group’. Finally, we use SKOS to
represent semantic relationships between each tag using properties such as
‘skos:broader’ and ‘skos:narrower’.
SCOT concepts and properties are formally defined to fully describe the tagging
model. There are core concepts called ‘scot:Tagcloud’ and ‘scot:Tag’ in the SCOT
ontology. ‘scot:Tagcloud’ is a class that can be used to represent information for the tag
cloud itself. The ‘Tag’ class, a member of the ‘tag cloud’ class, is identified to represent
the concept of a tag that has a name through URIs. All tags have an associated concept
and can be represented by a hierarchy (‘skos:broader’ and ‘skos:narrower’) among tags
using SKOS. It can provide a different structure to visualise a tag cloud beyond the flat
organisation of tags and will be an alternative way to overcome problems with such a flat
4 Wiki-based sociAl network thin client
4.1 Design principles
The proposed system aims to manage daily routine for knowledge workers more
effectively and efficiently. We have used personal digital assistants such as ‘to do’ lists,
calendars, address books, appointment books on both desktop and web environments.
The proposed system must provide integrated functionalities to support seamless
knowledge activities. In order to do this, the proposed system must support
interoperability among different platforms without loss of information and functionalities.
Also, there are two important design principles for developing the proposed system:
easy to combine online mash-up services: knowledge workers tend to do their
knowledge activities across desktop and web environments. For instance, they would
use online services such as del.icio.us or YouTube in order to manage and share their
knowledge. Therefore, the proposed system must get information from the web and
integrate it with local data or vice versa.
semantic tagging: as we mentioned in Section 3, tagging can be an effective method
to represent knowledge for knowledge workers – both at individual and community
level. But, a tag itself has limitations to describe the meaning of context in order to
precisely search resources. In addition, most wiki systems do not provide tag
features. The SCOT ontology can be helpful to describe semantic relationships
among tags. In order to support this, a proposed system should have a method to map
between tags and a SCOT ontology.
We now describe our prototype application called WANT. WANT [as extension to the
TiddlyWiki (http://tiddlywiki.com)] is a lightweight desktop wiki for PKM rather than
full-fledged semantic desktop applications. Figure 2 shows the architecture of WANT.
and JSON (interface layer), thus, allowing to easily extend with certain functionalities
content in the system are stored in HTML itself, and published using RSS formats and
SCOT ontology in the decentralised knowledge source layer. The mash-up services layer
plays an important role to connect desktop resources with web resources using semantic
web technologies: ‘connection to semantic desktop’. All contents can be saved in HTML
files, RSS1.0, RSS2.0 and Atom formats in WANT. There are two ways to connect the
semantic desktop components. Firstly, since WANT is a simple and single HTML file, it
could be opened by most semantic desktop applications. Secondly, all content of WANT
would be generated by RSS1.0 format and be saved in RDF storage. When querying with
RDF query languages such as SPARQL, users would be able to get the data set in the
RDF storage and return back as value the new semantic data.
Personal knowledge management for knowledge workers 35
Figure 2 WANT architecture (see online version for colours)
Connection to Web 2.0
Wikis are easy to use as collaboration platform and knowledge management systems.
However, using wikis will significantly limit the potential for information sharing and
collaboration in desktop environments. We overcome some of the weak points of wikis
by using Web 2.0 technologies. WANT allows knowledge workers to organise their
information or knowledge and provides various social content services such as
folksonomies, social bookmarking and RSS/Atom feeds using HTTP, SOAP, XML RPC
or REST web services. For instance, when knowledge workers make their own tags for
certain content, they can use not only desktop tags which they made before, but also a
folksonomy of weblogs or Technorati. All tags in the system can be translated and
mapped to a SCOT instance.
WANT uses browser-based interfaces. Figure 3 shows a sample content about the
picture ‘simplicity-desktop’ from Flickr as rendered in WANT. It can include pictures,
desktop resources and links.
H. Kim et al.
Figure 3 WANT interface: the picture ‘simplicity-desktop’ from Flickr (see online version for
Note: WANT includes the RSS reader, tag cloud and social bookmark reader.
5 How it works
5.1 Enabling social collaboration and authority
A user can create new content with a new link and modify existing contents in an editing
mode. All contents in the system could contain title, description, created date and tags.
These elements could enrich the metadata of the contents. The description allows a user
to organise his/her ideas or information about what they want to create. At the same time,
they can add some tags to an item in order to classify and categorise the content. The
system allows a user to assign multiple tags with separating comma like ‘Web 2.0,
desktop’. After creating content with tags, the tags can be located at the bottom of the
content and simultaneously the tags will be added in the tag cloud if the lists of tags are
not defined in the system. The tag cloud shows a list of popular tags. The tags can be later
used for navigating and finding contents. This can be a starting point to navigating
contents via tags.
One of the main goals of WANT is to foster and employ social interactions for
knowledge workers through various content services and Web 2.0 sites. To leverage the
social collaboration, we need to shift from focusing on the individuals to focusing on
interactions. Social collaboration within WANT is in particular supported by the RSS
reader, social bookmark reader and Flickr photo reader. Knowledge workers are able to
participate in communities sharing common interests and to have the links for references.
In WANT, it enables the users not only to capture web-based information, but also to
organise the information together with desktop resources. The collaboratively user-added
annotations are to improve social features on desktop resources. For instance, RSS reader
and social bookmark reader can get the data from a user-given URL such as a certain
RSS feed or bookmark URLs. The user can edit this data directly and add user-driven
annotations as tags in WANT. These tags can be connected with social communities as
specific links. So, contents on WANT easily reflect social content without the need for
the user’s additional effort.
Personal knowledge management for knowledge workers 37
Figure 4 RSS reader (1), social bookmark reader (2) and Flickr photo reader (3) (see online
version for colours)
5.2 Tag-based search
WANT provides two search methods: full-text search and tag-based search. The former
returns results for matching one or multiple keywords in content, while the simple
tag-based search navigates all tag lists in the tag cloud with the given tag. The tag cloud
is a list of the most popular tags used by the user. The larger the font size of the tag, the
greater its popularity. When the user navigates by a tag, he/she will be directly connected
with other resources using the tag cloud, which is clickable. It will improve the
H. Kim et al.
findability of content. If the user clicks on a certain tag, a list will pop up with links to the
other content having the same tag. This is supposed to give an indication of the most
popular topics. The availability of new semantic data will allow users to find and make
use of relevant data quickly and accurately.
Figure 5 Tag-based search (see online version for colours)
Note: When a user clicks a certain tag in the tag cloud, it shows a list of contents using
5.3 Enhancing semantics
There are several ways to store contents in the system:
1 static HTML files
2 RSS files
3 RDF repository.
Basically, all contents can be stored in HTML files which includes URLs, tags, contents,
links and date, etc. At the same time, all contents can be published with different
syndication types such as RSS1.0, RSS2.0 and ATOM.
It is a slightly different approach to the other semantic wikis. Most of those have a
way to generate semantic data in their tools directly. In our approach, we create contents
and syndication files, then we generate semantic data automatically to reuse and share it.
It is a more effective and efficient lightweight mechanism to connect with the semantic
desktop. In addition, if necessary, it can be integrated with other semantic desktop
applications because it is a simple HTML. It enables users to publish and share their
contents as RSS on the web. Also, if they want to collaborate with someone, they
exchange the RSS using an RSS reader. Thus, it is easy to cooperate in distributed
Personal knowledge management for knowledge workers 39
Example of RSS1.0
<rdf:RDF xmlns:rdf= “http://www.w3.org/1999/02/22–rdf–syntax–ns#”
<dc:title>Wiki based sociAl Network Thin client</dc:title>
<dc:description>The WANT is building....</dc:description>
<dc:title>How to Rip...</dc:title>
<dc:date>Sun, 1 Nov 2007 23:20:52 GMT</dc:date>
5.4 Mapping tags to SCOT ontology
Although RSS provides minimum functionality to describe semantically the contents in
the system, relationships among tags are still ambiguous. The SCOT ontology defines
each tag and its relationships at a semantic level, as well as frequencies for each tag.
Figure 6 presents a conceptual model for mapping tags to concepts in the SCOT
ontology. All tags define a concept as a ‘scot:Tag’ and can be represented by a
hierarchical structure among tags based on SKOS (Miles and Bechhofer, 2008). The
‘skos:broader’ property is used to describe a more general term and is the inverse of
‘skos:narrower’. There are two types of frequencies (i.e., absolute vs. relative) for
representing a tag frequency: ‘scot:AFrequency’ and ‘scot:RFrequency’. The former is
intended to describe an absolute value and the purpose of the latter is to represent the
relative value as a proportion of total occurrence. A single tag can have both frequency
formats. In particular, this information is represented by the ‘scot:ownAFrequency’ and
‘scot:ownRFrequency’ properties which are sub-properties of ‘scot:AFrequency’ and
‘scot:RFrequency’, respectively. Listing 2 displays the tags ‘blog’ and ‘web’. The tag
‘blog’ has 110 as an absolute frequency and 40.29% as a relative frequency. And it can
be seen that the tag ‘web’ is broader than the tag ‘blog’ and also there is a cooccurrence
relation between the two tags (i.e., ‘Cooccurrence_0’).
H. Kim et al.
A certain tag may often appear together with other tags. The meaning of the tag can
become more specific when the tag is combined with a set of tags and the frequency of
cooccurrence of two tags has more significance than the two individual ones. To simply
define the term ‘cooccurrence’: if an item contains both tags ‘blog’ and ‘web’, these tags
are said to cooccur or have a first order cooccurrence that is observed when tags cooccur
in the same items. It can play an important role in reducing ‘tag ambiguity’. When
describing cooccurring tags and their frequency among them in Figure 6, we need to
represent it by n-ary relations. The tag ‘blog’ has a cooccurrence with ‘web’ with AF and
RF of 5 and 0.4, respectively. The facts are represented using an instance of the class
‘scot:Cooccurrence’. The individual ‘Cooccurrence_0’ here represents a single object
encapsulating both the tag (web, a specific instance of ‘scot:Tag’) and
the cooccurrence value among the individuals (5). A tag cooccurrence is
represented by the ‘scot:cooccurTag’ property. The ‘scot:cooccurAFrequency’ and
‘scot:cooccurRFrequency’ property describe the frequency of the absolute and relative
cooccurrence amongst a set of tags (see Listing 3), respectively.
Listing 2 Representation tag ‘blog’ and ‘web’ in SCOT
<scot:Tag rdf: about= “http://www.example.com/want/tag/blog”>
<skos:narrower rdf:resource= “http://www.example.com/want/tag/web”/>
<scot:cooccurWith rdf:nodeID= “Cooccurrence_0”/>
<scot:Tag rdf:about= “http://www.example.com/want/tag/web”>
<skos:narrower rdf:resource= “http://www.example.com/want/tag/web”/>
<scot:cooccurWith rdf:nodeID= “Cooccurrence_0”/>
Personal knowledge management for knowledge workers 41
Conceptual model for mapping tags in a resource with concepts in the SCOT ontology
(see online version for colours)
Listing 3 Cooccurring relationships among two tags
<scot:Cooccurrence rdf:nodeID= “Cooccurrence_0”>
<scot:cooccurTag rdf:resource= “http://example.com/tag/web”/>
<scot:cooccurTag rdf:resource= “http://example.com/tag/blog”/>
6 Summary and conclusions
In this paper, we discussed the method for enhancing personal knowledge activities for
knowledge workers. In fact, traditional PKM applications intend to manage a specific
task such as a schedule or an address book. Knowledge workers have been collaborating
with their colleagues or community’s users across desktop and web environments.
Therefore, PKM systems must support social features for knowledge activities. It means
that this system is not only focused on managing data, but also on connecting people and
enabling them to share data between them.
42 Download full-text
H. Kim et al.
We suggested wiki-based content management and tag-based knowledge
representation to support personal knowledge activities and implemented WANT system.
The main achievement of this work, distinguishing our approach from existing
systems, is the establishment of an architecture to interact desktop and web, to create
metadata based on the SCOT ontology and to enrich desktop systems with social
intelligence. Our approach allows knowledge workers to interact and share their
resources among their desktop and social software more easily.
This study was supported by the Dankook University Post-Doc Grant in 2008, a grant of
the Science Foundation Ireland under Grant No. SFI/02/CE1/I131, the interoperable HER
Research and Development (A050909) and Ministry of Health and Welfare, Republic of
Breslin, G., Decker, S., Harth, A. and Bojars, U. (2006) ‘SIOC: an approach to connect web-based
communities’, The International Journal of Web-Based Communities, Vol. 2, No. 2.
Brickley, D. and Miles, A. (2005) SKOS Core Vocabulary Specification – W3C Working Draft 2,
available at http://www.w3.org/TR/2005/WD-swbp-skos-core-spec-20051102/.
Brickley, D. and Miller, L. (2005) FOAF Vocabulary Specification, available at
Campanini, S.E., Castagna, P. and Tazzoli, R. (2004) ‘Platypus wiki: a semantic wiki wiki web’, in
Proceedings of the 1st Italian Semantic Web Workshop Semantic Web Applications and
Perspectives (SWAP), Ancona, Italy.
Decker, S. and Frank, M. (2004) ‘The networked semantic desktop’, CEUR workshop proceedings,
in Proceedings of WWW Workshop on Application Design, Development and Implementation
Issues in the Semantic Web, New York, USA, available at
Golder, S. and Huberman, B.A. (2006) ‘Usage patterns of collaborative tagging systems’, Journal
of Information Science, Vol. 32, No. 2, pp.198–208.
Gruber, T. (2007) ‘Ontology of folksonomy: a mash-up of apples and oranges’, First On-line
Conference on Metadata and Semantics Research (MTSR ‘05), to be appear in International
Journal on Semantic Web and Information Systems, Vol. 3, No. 1.
Hendler, J. (2007) ‘The dark side of the semantic web’, IEEE Intelligent Systems, Vol. 22, No. 1,
Hepp, M., Siorpaes, K. and Bachlechner, D. (2007) ‘Harvesting wiki consensus: using wikipedia
entries as vocabulary for knowledge management’, IEEE Internet Computing, Vol. 11, No. 5,
Hotho, A., Jaschke, R., Schmitz, C. and Stumme, G. (2006) ‘Bibsonomy: a social bookmark and
publication sharing system’, Proceedings of the First Conceptual Structures Tool
Interoperability Workshop at the 14th International Conference on Conceptual Structures,
Aalborg Universitetsforlag, Aalborg, pp.87–102.
Kim, H.L., Breslin, G.J., Yang, S.K. and Kim, H.G. (2008) ‘Social semantic cloud of tag: semantic
model for social tagging’, in Proceedings of the 2nd KES International Symposium on Agent
and Multi-agent Systems: Technologies and Applications (Forthcoming), Incheon, Korea.
KM Magazine (2004) Personal Knowledge Management, Vol. 7, No. 7, available at