Kaspars Osis and Janis Grundspenkis
Institute of Applied Computer Systems, Riga Technical University, Meza 1/4, Riga LV 1048, Latvia
Keywords: Intelligent agents, Personal knowledge management, Knowledge worker.
Abstract: It is frequently mentioned that nowadays is the information age. Knowledge becomes the most important
asset for individuals and organizations. And more increasingly knowledge has been viewed as an active area
of research. Accordingly there is a need for highly qualified knowledge workers. That in turn implies a
necessity for on an effective technology based education system, which provides a foundation for obtaining
well educated specialists. Thus perspectives of personal knowledge management (PKM) environment are
explored in this context. This environment is not just focused on an individual. Rather it is involving also
collaboration for knowledge exchange thus forming communities of practice. The central concept of the
paper is knowledge worker surrounded by several layers of agents such as personal agents, communication
agents and so called “engine room” agents like database and network agents. The next step related to
different types of agents would be to consider that all or part of them could be mobile agents. Possible
future opportunities for PKM are explored in this respect and potential benefiting parties are identified.
In couple last decades one can observe a historic
transition from the industrial age to the information
age. Some may even argue that we already are in the
information age. Creation and consumption of
material goods, usage of fixed procedures and
following standardized information routines can be
named as characteristic elements of industrial age. In
opposite to latter mentioned, the information age can
be described as creating and consuming information,
using ad-hoc approaches and non-standardized
information for decision making and reaching
solutions. The Web and the Internet have shifted
even more focus towards the importance of
information. That changes the way information is
stored, presented, consumed and shared with others.
That in turn provides new options for doing business
in different areas - be it accounting, car engineering,
or teaching. The development of the Web has been
very rapid. So has been with the growth of
information amounts people and software systems
have to deal with. That is known as information
overload. In addition workspace equipment is
becoming more sophisticated which requires
additional skills and knowledge to handle it. As a
result work is becoming increasingly complex
(Wiig, 2004) and more complicated processing
systems have to be developed to make sense out of
these vast amounts of information. That has led to
recognition that knowledge has become the most
important asset for organizations and for individuals,
which more increasingly has been viewed as an
active area of research. This shift from information
to knowledge has been a reason for appearance of so
called “knowledge work”, which might be seen as a
new position within a list of intellectual jobs.
Increased importance and usage of knowledge in
business and in everyday life creates a necessity for
well educated people. That implies a need for on an
effective technology based knowledge management
and education system, which would pave the road
for a new knowledge-based society and economy,
and would allow ambient participation in a social
and economic life. This paper reports on one stage
of broader research targeted at perspectives of using
agents in the environment of personal knowledge
management. At this stage of research a personal-
knowledge based worker environment supported by
number of agents is defined. And next it is viewed in
perspective of PKM and possible future
Osis K. and Grundspenkis J. (2009).
In Proceedings of the International Conference on Knowledge Management and Information Sharing, pages 332-337
DOI: 10.5220/0002331303320337
One could assume that over twenty years it is more
than enough to get straight with the definition of the
basic concepts. This is not the case regarding agents.
The concept ‘agent’ has not been defined in a single
unified way which could be widely accepted. Thus
there are several approaches how to define an agent.
The term “agent” itself surfaced for the first time in
mid-1950s with John McCarthy (Bradshaw, 1997).
The American Heritage Dictionary defines an agent
as “one that acts or has the power or authority to act;
or one empowered to act for or represent another; or
a means by which something is done or caused;
instrument” (Pickett, 2000). General software agent
definition says that an agent is every program that
acts in the name of its user (i.e. human) (Bradshaw,
1997). A more specific definition of a software agent
which could be more widely acceptable is given by
Shoham: a software entity which functions
continuously and autonomously in a particular
environment, often inhabited by other agents and
processes (Bradshaw, 1997). Also Wooldridge and
Jennings point out that first of all an agent is a
computer system situated in some environment, and
that it is capable of autonomous action in this
environment in order to meet its design objectives
(Wooldridge and Jennings, 1995). Agents’
autonomy is perceived as equivalent ability to
humans’ free will (Vidal and Buhler, et. al. 2001).
Agents may or may not have several
characteristics or features. They are: autonomy,
social ability, reactivity, pro-activity, mobility,
veracity, benevolence, rationality (Wooldridge and
Jennings, 1995; Padgham and Winikoff, et. al.
2008). In addition to already mentioned agent
features Bradshaw adds: temporal continuity,
personality, adaptivity (Bradshaw, 1997).
Agents can be divided in types. Nwana comes up
with several of them. He classifies agents as
collaborative agents, interface agents, mobile agents,
information agents, reactive agents, hybrid agents,
heterogeneous agent systems, and smart or
intelligent agents (Nwana 1996).
2.1 Intelligent Agents
A bit more about intelligent agents - what are they?
Wooldridge defines intelligent agents as ones that
are forced to work sturdy in rapidly changing,
unpredictable, and open environments, where exists
a good possibility that actions may fail (Wooldridge,
1999). He adds that an agent is a computer system
capable of flexible autonomous action in order to
meet its design objectives. By flexible he means that
such system must be responsive, proactive and
social. Australian software company JACK
intelligent agents in their commercial applications.
There an intelligent agent is being seen as an
encapsulated computer system that works within
larger systems, or in other environments, and one
that can reason. It can perform functions that require
higher-level cognitive abilities (AOS Group, 2009).
2.2 Mobile Agents
In turn a mobile agent is an execution unit able to
migrate in an autonomic way to another host,
transporting along with itself its code and state; and
seamlessly resume its execution in this new
environment before that installing its own code.
(Nwana, 1996; Lange and Oshima, 1999). The term
“state” usually means agent parameters’ values,
which helps agents to realize from which place to
continue to execute interrupted work; “code” in
other hand in a sense of object oriented context
means class code, used by agent to execute itself
(Lange and Oshima, 1999).
The term ‘personal knowledge management’
consists of three words – personal, knowledge and
management. Obviously personal refers to an
individual and to everyday tasks he or she performs
or is intending to do. What is knowledge? If to look
in a personal – human perspective, then knowledge
and its physical location is mainly in the brain and
thus it appears as our memories and skills
(Apshvalka, 2004). However she also points out that
knowledge is rather intangible and that it can not be
fully realizable in common with all our human
being. Thus considerable role is played by
individual’s characteristics such as volition,
psychological traits, motivation and his or her
intelligence. A broader definition is nailed down by
Thomas Davenport, where knowledge is “a fluid
mix of framed experience, values, contextual
information, and expert insight that provides a
framework for evaluating and incorporating new
experiences and information. It originates and is
applied in the minds of knowers. In organizations it
often becomes embedded not only in documents or
repositories but also in organizational routines,
processes, practices, and norms” (Davenport and
Prusak, 2000). Knowledge management (KM) in its
classic way is viewed in business context in
organization. Thus KM is defined as such which
enables creation, sharing and utilization of
knowledge in order to achieve business goals
(Quintas 1999). The primary goal of KM in this
context is identified as “to facilitate opportunistic
application of fragmented knowledge through
integration” (Tiwana 2002).
PKM is rather overlooked area within KM even
though lately it picks up speed again. “Definitions of
PKM revolve around a set of core issues: managing
and supporting personal knowledge and information
so that it is accessible, meaningful and valuable to
the individual; maintaining networks, contacts and
communities; making life easier and more
enjoyable; and exploiting personal capital”
(Higgison 2004). Eric Tsui defines PKM as “a
collection of processes that an individual needs to
carry out in order to gather, classify, store, search
and retrieve knowledge in his/her daily activities”.
He adds that “activities are not confined to
business/work-related tasks but also include personal
interests, hobbies, home, family and leisure
activities” (Tsui 2002). More laconic is Steve Barth
as he says that PKM is taking responsibility for what
you know, who you know, and what they know
(Barth 2000). Within essence of this short definition
lays the cultural and collaborative aspect of PKM.
Meaning that PKM is not focused just on an
individual, but it is more concentrating on culture
and collaboration between knowledge workers. Thus
PKM is fostering a creation of communities of
practice (CoPs) which serve as a fertile ground for
knowledge sharing and subsequently for knowledge
creation. There are already a number of different
types of CoPs established online starting from
professional, business, everyday practical questions
related ones and ending with social network type of
CoPs. Users’ comments next to bidder’s history on
eBay auction site or customers’ descriptions of
bought goods on are examples of
CoPs related to everyday practical questions. Instead
LinkedIn is both an example of a professional CoP
and a social network CoP.
3.1 Knowledge Worker
Thus PKM is also geared toward CoPs, culture and
collaboration of knowledge workers. That requests
for a closer look at these individuals. Who are they?
For the first time the term “knowledge worker”
appears in a book by Peter Drucker (Drucker 1959)
in the middle of previous century. Since then this
term has been looked at several times. Really in the
spot light it started to appear at the end of previous
century. A well known author on knowledge and
knowledge management Thomas Davenport defines
knowledge workers as ones that have high degrees
of expertise, education, or experience, and the
primary purpose for their jobs involves the creation,
distribution, or application of knowledge (Davenport
2005). In short, intellectual job they do is the way of
working they do for living. Davenport goes even
further – he names categories or areas where
knowledge workers would be most probably located.
These are: management; business and financial
operations; computers and mathematics; architecture
and engineering; life, physical and social scientists;
legal area; healthcare practitioners; community and
social services; education, training and library; and
arts, design, entertainment, sports, media (Davenport
2005). He adds, that this forms a respectable almost
one third of all the labor force in United States. No
doubts that similar situation might be observed in
other countries as well.
If to bring focus back on the individual himself
or herself, then ideally, knowledge workers should
possess not just technical know-how, but also sure
sense of the cultural, political, and personal aspects
of knowledge (Davenport 2000). That means that
personality characteristics of knowledge workers
most probably play an important role in how he or
she is at finding, understanding, and making use of
organizational knowledge (Dalkir 2005). This is true
not just in case of organizational knowledge. It is
true also in managing and enhancing knowledge
worker’s own knowledge as well. John Brown says
that innovation takes place at all levels of the
company – not just in the research department
(Drucker and Garvin, et. al. 1998). This can be
derived even further by saying that innovation and
new ideas in knowledge worker’s level happen at all
places and times within his or her daily routine – not
just at work or school. Thus it is important that
knowledge worker has at hand a PKM system
(PKMS). It is a complex system and includes
psychological, social and technological aspects
(Apshvalka and Grundspenkis, 2005). The
performance of PKMS is conditioned with
knowledge workers emotions, perceptions, believes,
objectives, surrounding society and environment.
Also technologies play an important role. Such
system can serve as a support for performing simple
information management tasks as well as a support
for much more intellectual activities, for example,
such as collaborative learning of a new language
while commuting by train.
KMIS 2009 - International Conference on Knowledge Management and Information Sharing
Figure 1: Knowledge worker’s agent environment.
If to look in more details one can easily see that an
individual might be involved in a number of very
different activities that requires knowledge and
skills. He or she might be studying at university a
new subject, looking for a new or better job, setting
up a meeting with peers to get help on chemistry
homework, booking a plain ticket for attending an
international conference and so on. Many times in
these activities a searching is involved for
information, communication, scheduling or
messaging just to mention a few of them. In PKMS
all these activities require an intelligent support,
which may be implemented in the form of
communities of intelligent agents. There can be
distinguished three groups of agents (see Figure 1)
that could form a basis for knowledge worker’s
agent environment.
First, agents that can serve as the hard work
performers – a driving force for so called “engine-
room”, which basically is an integrated set of
hardware, software and technologies to assure
knowledge acquisition, processing, storage and
representation as it is stated for organizations in
(Grundspenkis, 2003). Here we try to use this
concept for PKM. Secondly, they are agents that
enable communications. And finally, these are
personal agents that are most closely tied with
knowledge workers.
The idea behind “engine room” agents is that
nowadays there are plenty things to-do related with
rather technical tasks that knowledge worker has to
embrace in order to streamline his or her daily
activities. Such agents, for example, could provide a
helpful hand in monitoring and collecting
information from data streams. They could
appropriately react on changes in combined “engine
room” environment, which could consist from LAN,
WAN, the Internet, hardware used by knowledge
worker, etc. Such environment for most people is
increasingly difficult and time consuming to handle.
As “engine room” agents can be mentioned database
agents – for storing knowledge elements and
information, hardware agents – for performing
system adjustments to particular user’s hardware,
physical network agents – for supporting PKMS on
a network level, connection agents – for establishing
connections and determining appropriate protocols,
local content agents – for taking care of different
knowledge elements stored on user’s device,
intelligent Web agents – for performing advanced
knowledge elements’ acquiring and processing
activities on the Web, and social network agents –
for supporting low level activities in relation with
social networks. Also there can be mentioned other
agents that provide means to support technologies
for fundamental functions of knowledge work, for
example, software distribution agents, which main
goal is to take care of timely software updates and
software installation package deliveries.
Other group of intelligent agents is
communication agents. As their name says by itself,
these agents are in charge of communications. For
individual situated in multi-agent environment
communications is very important point of focus in
order to have effective knowledge creation,
acquisition, sharing and distribution in the PKMS.
These agents include messaging agents (Knapik and
Johnson, 1998), cooperative agents that are geared
towards communication with other agents and
collaborative agents that focus on performing tasks
in collaboration with other agents. Thus these agents
support the collaboration, communication and
culture characteristics of PKM. Also team agents
(Ellis and Wainer, 2002) that support groupware
technologies can be mentioned as part of
communication agents.
Personal agents are ones that are directly
influenced by the individual, support interaction
with particular hardware device, and provide help in
knowledge work. These agents include assistant
agents, search agents, filtering agents and workflow
agents (Knapik and Johnson, 1998). Assistant agents
can provide help with automated hotel or flight
booking, scheduling meetings by taking into account
available time frames and locations, or they could
handle individual’s e-mail system by sorting
incoming e-mails and reminding to reply. Search
agents are most well know and widely used ones, for
example, one can think about Paper-clip agent in MS
Office. Such agents can be geared towards different
type of searches like keywords in scientific articles
or database indexes, or yellow page directories on
corporate network. Filtering agents can be used for
monitoring information on the Web or for suspicious
pattern recognition in home video surveillance
system. Workflow agents can be helpful in
scheduling daily tasks. This list of agents could be
expanded by adding efficiency agents that help to
coordinate, prioritize and schedule other personal
agents work in order to achieve higher performance
results. That is especially important in perspective of
mobile devices used as the basic hardware platform
for PKMS as they have limited resources and
connection time with the Internet or with other type
of network.
The era of intelligent agents approaches quite
fast even though still are being used rather simple
agents. However despite usage of these relatively
simple agents there are quite good options to
develop an agent-supported environment for
knowledge worker in the PKMS. In Figure 1 is
depicted proposed knowledge worker’s agent
environment. Solid directed lines denote data,
information and knowledge flows. Doted directed
lines denote possible connections between involved
Nowadays are being done first movements towards
online community where intelligent agents will
support individuals in their knowledge work by
helping to get a better grasp on knowledge and
resources located out there on the Web. In particular
the Semantic Web is considered to be the way to go
as the regular Web is missing common uniform data
and information encoding language and common
ontology to represent knowledge. Thus it is
understandable to humans, but not to agents. On
contrary the Semantic Web is the right environment
for intelligent agents, which enables them to help
individuals in better management of their personal
knowledge. Agents will team up in groups to
achieve better performance by working as one whole
unit thus outperforming single intelligent agents.
Another future perspective lays in the idea that
part of agents in the proposed knowledge worker’s
environment could be mobile agents. That would be
especially beneficial in case of using mobile devices.
They have limited memory and processing resources
as well as connections to network are unstable,
KMIS 2009 - International Conference on Knowledge Management and Information Sharing
which is very true in remote areas. Thus individual,
for example, could dispatch agent with a task and
reconnect back to network just after a week to find
out the status.
In either way proposed knowledge worker’s
agent-based environment is being seen as a source
for potential direction of creating more enhanced
PKMS. It is also believed to serve as a foundation
for researchers to perform investigation in different
domains like mobile learning, personal business or
healthcare where proposed environment in context
of personal knowledge management approach could
be beneficial to apply.
AOS Group, 2009. Agent Technology, What are Software
Agents? Retrieved May 30
from, http://www.agent-
Apshvalka, D., 2004. Personal Knowledge Management.
In ECITE’04 Proceedings of the 11th European
Conference on Information Technology Evaluation.
Amsterdam, Netherlands, 17–22.
Apshvalka, D., Grundspenkis, J., 2005. Personal and
Organizational Knowledge Management as a Driving
Force for Business Process Effectiveness. In
Proceedings of the International Conference on
Information Technologies for Business. Vilnius
University, Kaunas, 9–14.
Barth, S., 2000. The Power of One, Knowledge
Management Magazine. Retrieved May 30
Bradshaw, J. M., 1997. Software Agents, AAAI Press. 2
Dalkir, K., 2005. Knowledge Management in Theory and
Practice, Butterworth-Heinemann. Oxford.
Davenport, T.H., Prusak, L., 2000. Working knowledge:
how organizations manage what they know, Harvard
Business School Press, Boston, Massachusetts.
Davenport, T.H., 2005. Thinking for a living: how to get
better performance and results from knowledge
workers, Harvard Business School Press. Boston,
Drucker, P.F., 1959. The Landmarks of Tomorrow,
Harper. New York.
Drucker, P.F., Garvin, D., Leonard, D., 1998. Harvard
business review on knowledge management, Harvard
Business School Press. Boston, Massachusetts.
Ellis, C., Wainer, J., 2002. Groupware and Computer
Supported Cooperative Work, In: Multiagent Systems,
A Modern Approach to Distributed Artificial
Intelligence (Waiss, G., Ed.). MIT Press,
Massachusetts, 425-458.
Grundspenkis, J., 2003. Development of Hybrid Intelligent
Systems: Integration of Structural Modeling,
Intelligent Agent and Knowledge Management
Techniques. In Scientific Proceedings of Riga
Technical University, 5th series, Computer Science,
Applied Computer Systems. RTU Publishing, Riga.
Vol. 17, 7-30.
Higgison, S., 2004. “Your say: personal knowledge
management”, Inside Knowledge, 7(7). Retrieved May
Knapik, M., Johnson, J., 1998. Developing intelligent
agents for distributed systems: exploring architecture,
technologies, and applications, McGraw-Hill. New
Lange, D., Oshima, M., 1999. Seven good reasons for
mobile agents, Communications of the ACM. March
42(3), 88-89.
Nwana, H., 1996. Software Agents: An Overview.
Knowledge Engineering Review, Cambridge
University Press, 11(3) September, 1–40.
Padgham, L., Winikoff, M., DeLoach, S., 2008. A Unified
Graphical Notation for AOSE. In AOSE’08, 9th
International Workshop on Agent-Oriented Software
Engineering at AAMAS’08 The 7th International Joint
Conference on Autonomous Agents and Multi-Agent
Systems. Estoril, Portugal, 1-12.
Pickett J. P., 2000. The American Heritage Dictionary of
the English Language, Houghton Mifflin. 4
Quintas, P., 1999. Quoted in: Open Eye - Head Back to
the Business Cafe, The Independent, London,
February 4, Retrieved May 30
Tiwana, A., 2002. The Knowledge Management Toolkit:
Orchestrating It, Strategy, and Knowledge Platforms,
Prentice Hall. Upper Saddle River.
Tsui, E., 2002. Technologies for Personal and Peer-to-Peer
(P2P) Knowledge Management, CSC Leading Edge
Forum, Technical Report. Australia.
Vidal, J., Buhler, P., Huhns, M., 2001. Inside an Agent.
IEEE Internet Computing, February, 82–86.
Wiig, K. M., 2004. People-focused knowledge
management: how effective decision making leads to
corporate success, Butterworth-Heinemann. Oxford.
Wooldridge, M., 1999. Intelligent Agents, In: Multiagent
Systems, (Waiss, G., Ed.), MIT Press. Massachusetts,
Wooldridge, M., Jennings, N. R., 1995. Intelligent agents:
Theory and practice. The Knowledge Engineering
Review, 10(2), 115–152.