Personal knowledge management (PKM) has come to be a central and critical element in the spectrum of knowledge management. The context and control of knowledge management is shifting from global and organizational levels to communities and ultimately to individuals. In developing a model for managing personal intellectual capital, two basic issues stand out. First, we must attend to the broad context of the individual represented on social networks, collaborative environments and other channels. Second, individuals should assume ownership, active management and control over information about themselves by building a formal PKM profile. We can collect elements of our personal information, aided by semantic and knowledge technologies for discovery and normalization, to compile and manage as our own intellectual capital. Work in progress at Kent State University has highlighted important lessons in personal knowledge management, including the foundational role of semantics and the need for a standard definition for a personal data model.

personal information
knowledge management
information resources management
data models
semantics
knowledge discovery

Bulletin, December/January 2012


Enabling Personal Knowledge Management with Collaborative and Semantic Technologies 

by Denise A. D. Bedford

In this article knowledge management is seen from the personal perspective. It is about people, what they know, how they learn and how they innovate. In 2011, four trends are redefining our understanding of knowledge management and highlighting the importance of both strategic organizational and personal knowledge management (PKM). They are: 

  1. The shift from an industrial to a knowledge economy 
  2. The rapid development of semantic and knowledge technologies – covering both Web 2.0 and Web 3.0 
  3. The virtualization of work and the working environment
  4. A fundamental shift from application-based, information-centric architectures and technologies to open and knowledge-centric architectures and technologies. 

The shift from an industrial to a knowledge economy positions intellectual capital [1, 2, 3, 4, 5] as the central factor of economic growth and production. In a knowledge economy, knowledge powers transactions, production and consumption. In the knowledge economy, there are heightened incentives to create, share, mobilize and preserve all forms of knowledge. Open knowledge markets support knowledge transactions and allow knowledge producers and consumers to establish the value of knowledge based on a particular context. It is the individual person that engages in knowledge exchanges and transactions, learns, acquires new knowledge and validates or invalidates knowledge. The primary agents in a knowledge economy are the individuals who possess the intellectual capital. 

The rapid development of semantic and knowledge technologies – including those that power Web 2.0 and Web 3.0 – is the second trend. Web 2.0 focuses on interaction, the engagement of people and communities. Web 2.0 provides the context in which knowledge transactions occur. Web 2.0 supports unobtrusive capture of semi-tacit knowledge or knowledge that is un-reviewed, serendipitous and closer to raw. The new collaborative and social environments make it much easier to discover and capture all forms of knowledge at an earlier point in the knowledge life cycle [6, 7]. 

According to Mills Davis [8], a high-profile advocate of the semantic web, Web 3.0 is about semantics and semantic technologies. Semantic technologies have the potential to expand the scale and scope of knowledge transactions. Semantic and knowledge technologies enable us to capture and leverage more knowledge at a faster pace and to leverage the capacity of machine-based agents. Semantic and knowledge technologies provide the tools and opportunity to encode human knowledge so that it is available to a broader population. 

The third trend is the virtualization of the working environment and the increasing popularity of remote working [9, 10]. People will work where and when they want to work. Workers will do their work in an environment that is designed for them. Workers will have personalized agents that find information for them regardless of where it lives. The new virtual environment supports dynamic people-to-people connections. It leverages and builds upon the two previous trends to support collaboration and facilitate the development of knowledge-centric cultures. Much of this new work environment exists beyond any one institution’s information management or information technology infrastructure. In this new virtual working environment, knowledge must flow beyond the application in which it was originally created or stored. Furthermore, knowledge is made accessible to other environments and applications. 

A fourth trend is a fundamental shift from traditional information architectures and information management technologies to knowledge architectures and knowledge management technologies [11]. The challenge this trend presents is as profound and fundamental as the challenges referenced above. Our current infrastructure is solidly grounded in the information management technologies that were developed from the 1980s through the early 2000s. These technologies served us well during those decades but they represent a pre-semantic and pre-knowledge mindset where information is packaged, stored and locked down, rather than a viewpoint in which knowledge is dynamic, continuously evolving and free flowing. Knowledge becomes a complex object. In this context, it is clear that people are a knowledge object. Our challenge is to determine how to represent people as knowledge objects and to understand how their knowledge may be managed, maintained, accessed, mobilized, exposed for consumption and protected. 

The Traditional and the Future Focus of Knowledge Management
Over the past 20 years, we have made significant advances in our understanding of the theory and practice of knowledge management. Traditionally, we began by looking at knowledge management from a world level, at the levels of national economies. In the past decade we have expanded our understanding of community and group knowledge. Today, forces such as the focus on intellectual capital, the virtualization of the working environment, semantic technologies and the shift to knowledge architectures focus our attention and efforts on personal knowledge – at the individual level. 

Figure 1
Figure 1. Shifting contexts – world, organizational, community and individual

Two Fundamental Questions
This shift raises two fundamental questions. Both questions can be simply stated but they have profound implications. Question one: If people are the primary source of knowledge in 2010, what is the fundamental representation of an individual’s knowledge? What constitutes metadata and meta-information about people? The information science discipline offers decades of experience and advice on metadata and bibliographic profiles for information objects and information packages. Can that experience be applied to defining descriptive representations of people as knowledge objects? To answer this question, we need to understand all of the dimensions of a person as a knowledge object. In order to operate in the new working environment described above, we need people profiles that are similar to but more extensible and flexible than metadata profiles for books or reports. People are more complex than books and reports. 

The second question arises from the first: If there is a universal people profile and data model, who owns it, where does it live and how do we maintain it? The issue is that entities other than the individual will create and promote information about the individual. The four trends discussed earlier make it easier to access information about an individual and to create an image for an individual. While there is little we can do to stop these trends, we can create a counterbalance. The counterbalance involves helping individuals take responsibility for their own profile and identity and their own knowledge representation and to manage that representation in the four contexts (see Figure 1). Individuals need to understand how to create a personal knowledge profile and how to actively manage it so that what is accessible is appropriate to the context. 

Figure 2
Figure 2. Shift from organizational to community to individual knowledge

Question Two: Raising Awareness of Context
Reversing the order of the questions for the moment, let’s assume that we have a universal people profile. Who creates the data that fills this profile? Where does it live? Who owns it? How is it maintained? And who can access it? To answer this question, we need to understand the context in which we would be using people profiles and data models. The answer depends on whether you are in a closed personal space, a community space, an organizational space or an open world environment. We suggest that the knowledge contexts identified in Figure 2 can help us to define our behaviors, set our expectations for management and access and better understand what is expected of the governance model. 

The individual space is accessible to and governed by a person. In this space we would expect to be able to represent all facets of our intellectual capital. We would also expect to be able to determine who can see which facets and under what conditions. In the future, the extended capabilities that are available in the individual space will be protected and firewalled just as an organization is firewalled today. Today the capability to create a profile that represents all of our knowledge, all of our intellectual assets, does not exist. The data that would comprise this representation is scattered across applications that we do not control. Such capabilities are not only possible but are also inevitable, given the four trends discussed earlier. The critical question, as Dede [12] suggests, is this: Can we create a counterbalance that puts control and ownership of an individual knowledge representation in the hands of the individual? This is the challenge for the next 10years. 

The community context exists today in the form of collaborative environments, social networks, community spaces and all forms of communication channels. The community context provides rich opportunities for an individual to represent them – to express ideas, thought and raw knowledge. This context allows us to capture those ideas and raw knowledge in a persistent way. What we now express in a social context is publicly and persistently accessible. In this context, we expect to engage in shared transactions. We understand implicitly that we share control with the others who are part of that community, as well as whoever owns the platform in which those transactions are taking place. The governance rules and protocols are generally known and in many cases may be defined by the members of the community. We understand what should be shared and what shouldn’t be shared in this space. We don’t make all of our intellectual capital assets available in this context. 

The organizational context is similar to the community context in that we expect shared transactions. We also expect more formal organizational controls and less opportunity for individuals to represent themselves in ways that are not sanctioned by the organizations. 

Third party transactions make up the world context. While we may be participants in those transactions, how we are represented is largely controlled by others. Our ability to influence or counterbalance those representations is limited. 

Question One: Representing the Intellectual Capital of an Individual
Returning to Question 1, we observe that today, the need for individuals to be aware of and manage their own intellectual capital and to control their own knowledge representation aligns with an organizational need to manage intellectual capital assets. Over the past six decades, knowledge management theory and practice have come to understand that the individual is the primary source of knowledge. Organizations must strategically manage and value their knowledge and intellectual assets just as they value their financial and physical assets. Organizations need to ensure that their intellectual capital assets grow, are mobilized and leveraged to create new economic growth. Ultimately, though, only an individual can invest in, grow and mobilize knowledge and create intellectual capital. What do we mean by intellectual capital? How do we gauge the intellectual capital of an individual? And what can we do to help individuals grow their intellectual capital? 

A classic definition of intellectual capital provided by Daniel Andriessen [13], illustrated in Figure 3, places intellectual capital in the context of two other types of capital – tangible capital and financial capital. Tangible capital includes land, fixed resources, plants, equipment and those resources that were critical to both the agricultural and the industrial economies. Financial capital is self-explanatory – it was a critical factor of production and growth for the industrial economy. Intellectual capital, a key factor of the knowledge economy, is defined to include human capital, structural capital and relational capital. Human capital includes implicit knowledge, skills and attitude. Andriessen defines structural capital to include explicit, encoded knowledge, processes and procedural know-how, and all forms of culture, including organizational, personal and national. Relational capital includes reputational knowledge and relational knowledge such as an individual’s networks, social relationships and business relationships.

Figure 3
Figure 3. Andriessen’s characterization of intellectual capital management

Over the past 60 years, we shifted our perspective from a national accounting perspective to an individual worker perspective. In 2011, we need to translate this high level definition of intellectual capital management to a personal knowledge management (PKM) space. In 2011 it is important that we take this concrete step forward. The Kent State University IAKM (information architecture and knowledge management) program is launching a pilot test to determine what such a personal intellectual capital/knowledge management model would look like. Let’s consider what such a PKM conceptual model might resemble. 

From Intellectual Capital to PKM 
For the purpose of this discussion, we adopt a broad definition of PKM that includes the ideas expressed by Davenport [14], Frand and Hixson [15], Pauleen [16], Grundspenkis [17] and Wright [18]. Essentially, our conceptualization of PKM is as multi-faceted as the domain of knowledge management itself. PKM simply shifts the focus of key knowledge competencies from the strategic and organizational level to an individual level. PKM refers to the individual’s competencies in knowledge leadership, knowledge culture, collaboration and communities, knowledge asset management, personal knowledge architectures and personal learning. What does a PKM conceptual model look like? 

Figure 4
Figure 4. Personal level translation of forms of intellectual capital

We can begin with Andriessen’s intellectual capital model. Each facet of Andriessen’s model can be translated to a PKM characteristic. Figure 4 provides a high level representation of personal knowledge and intellectual capital facets. Each facet of Andriessen’s intellectual capital model can be translated into a rich set of PKM sources and indicators. 

Designing a Dynamic Personal KM Profile 
The PKM conceptual model provides us with a framework for designing a PKM profile. A PKM profile can be operationalized as a dashboard for an individual, who could then use it to monitor, manage and present their own intellectual capital. 

Figure 5
Figure 5  Translation of human capital attributes to PKM characteristics

Figure 6
Figure 6. Translation of structural capital attributes to PKM characteristics

The first step in constructing a PKM profile involves a formal translation of the intellectual capital facets of Andriessen’s model into personal knowledge behaviors and characteristics. A possible translation of human capital factors to PKM attributes is illustrated in Figure 5. A possible translation of structural capital to PKM is illustrated in Figure 6. 

Figure 7
Figure 7. Alignment of intellectual capital factors, indicators and sources of evidence operationalizing the PKM model

The second step in building a PKM profile is data collection and evidence discovery – finding the critical sources and examples of personal knowledge to use to build the profile. Figure 7 provides examples of sources of evidence that might support the capture of PKM indicators. As we can see from some of the translation examples in Figure 7, such data collection is not a trivial task. Prior to the advent of Web 2.0 and the availability of collaborative and social media, this task would have been daunting to undertake once and almost impossible to consider maintaining on a continuous basis. However, with the availability and use of Web 2.0 technologies, finding and accessing these sources of data and evidence is possible. 

The purpose of defining PKM profiles is to provide individuals with the capability to monitor their own intellectual capital creation, growth and use and to publish PKM profiles to their communities, their organizations and to the broader world. Operationalizing a personal management profile can be a labor-intensive and subjective undertaking. In order to operationalize the PKM profile, we need to leverage semantic and knowledge technologies – Web 3.0 capabilities. 

At this stage we encounter two major challenges: (1) the varied form and nature of the evidence and its scatter across many applications and (2) the inherently subjective nature of some of our proposed indicators. For example, let’s consider how we would define indicators of an individual’s procedural knowledge. An individual might want to include in her profile the performance appraisal feedback she had received. Such feedback is generally embedded in other applications and wrapped in security. From organizational unit to organizational unit and across organizations the form in which this information is captured may vary widely. And, by definition, the performance appraisal feedback is subjective – provided by individuals and interpreted by individuals. A second example of procedural knowledge might consist of business decisions that an individual has made or ways of doing business that a person has defined. Again, this information may be embedded in business applications as business rules in an enterprise resource management application. A third example of procedural knowledge may be business process knowledge that an individual has created as part of fulfilling his work obligations – it may be represented as unstructured business documents that are linked to a business process or even represented as business knowledge in an email communication. Discovering this evidence will be challenging and, again, interpreting it may be subjective. 

Semantic and knowledge technologies such as those referenced earlier in Mills Davis’ roadmap [8] provide the tools that allow us to discover evidence and to interpret it in order to develop PKM profiles. For example, sophisticated semantic technologies, when configured with relevant language and characterization about business processes, can assist us in discovering and interpreting evidence across applications. Interpretation of evidence as indicators can also be achieved using semantic analysis technologies where the technologies have strong natural language processing foundations and where there is a capability to engineer human knowledge into the application. 

Sample Use Case 
Let’s take as a use case the interpretation of business decisions or business communication as an indicator of procedural knowledge. Let’s assume that we have examples of business tasks and decisions represented as decision explanations in business forms or as narrative in email communications. Our goal is to interpret this feedback to determine something about the nature and depth of the individual’s procedural knowledge. Let’s also assume that we have a sophisticated semantic analysis technology to work with. 

Our first step will be to determine what is important to us in terms of business knowledge – what is the nature of the language used? How complex is this language? How pertinent is the knowledge this person has expressed to the technical competencies we expect for this business task? Is there an indication of growth or learning in these competencies over time? Is the language used sufficiently expressive that others might learn from reading it? All of these parameters can be characterized through knowledge engineering methods and encoded as conceptual rules and parameters in semantic analysis technologies. 

The next step is then to select a sample set of individuals for pilot testing. The pilot test should take place over two years, beginning in Fall 2011. The sample set is being identified at this time. Individuals who are interested in participating or who would like to recommend others for participation or profiling are encouraged to contact the project team. 

Observations and Lessons Learned
We have been exploring these issues for just one short year now. Our early exploration, though, suggests that it is possible to configure and design semantic applications that will support the construction of PKM profiles. As with all early stages of new applications, however, the scale and scope of development is not yet cost effective. Those areas of the profile that are specific to structural capital in particular – where the embedded knowledge is closely aligned with a particular business domain or a way of working at an organization – may require targeted design and development. Nonetheless, other areas of the profile such as those aligned with human capital or relational capital may have a broader, cross-organization and cross-sector appeal. 

We have learned several lessons in exploring these ideas over the past year. A central lesson is that the most important knowledge for building semantic profiles to support PKM is not knowledge of the technologies but rather the semantics of business and procedural knowledge. Constructing a semantic application that is intended to profile procedural knowledge must be grounded on the semantics of procedural knowledge. Similarly a semantic application that would profile narrative intelligence must be grounded on the semantics of narrative intelligence. 

A second key lesson learned is that the core component of a PKM foundation is lacking – a standard definition for a people data-model. The expanded focus of knowledge management in the 21st century to include people and knowledge itself will require that we fill this gap. At this point in time, the only instance we found of even the beginning of a standard people data-model was in fact in social media applications. It is important for knowledge professionals to become more involved in the definition, description and standardization of people data-models and profiles. 

A third key lesson is that while the intellectual capital models provide a strong foundation for representing an individual, they are not complete. Individuals possess other assets that should be included in a universal people profile. 

A fourth key lesson learned is that semantic technologies that support the configuration and embedding of human knowledge or knowledge organization systems are likely to be the most productive for supporting this effort. While the semantic technology market is vibrant and rich with tools, there are some key components that appear to be more productive for this type of knowledge-management-focused work than others. 

Resources Mentioned in the Article
[1] Machlup, F. (1962). The production and distribution of knowledge in the United States. Princeton, NJ: Princeton University Press. 

[2] Porat, M. U. (1977). The information economy (Vols. 1-9). Washington, DC: Office of Telecommunication, United States Department of Commerce. 

[3] Foray, D. (2004). Economics of knowledge. Cambridge, MA: MIT Press. 

[4] Al-Ali, N. (2003). Comprehensive intellectual capital management: Step-by-step. New York: Wiley. 

[5] Allee, V. (1997). The knowledge evolution: Expanding organizational intelligence. Newton, MA: Butterworth-Heinemann. 

[6] Levy, M. (2009). WEB 2.0 implications on knowledge management. Journal of Knowledge Management, 13(1), 120 -134.

[7] Chatti, M.A., Klamma, R., Jarke, M., & Naeve, A. (2007) The Web 2.0 driven SECI model based learning process. In Seventh IEEE International Conference on Advanced Learning Technologies – ICALT 2007 [pp. 780-782]. Los Alamitos, CA: IEEE Computer Society. 

[8] Semantic Wave. (2008). Industry roadmap to web 3.0 & multibillion dollar market opportunities. (Semantic Wave Report). Washington, DC: Project 10X. 

[9] Stewart, J., & R. Williams (1998). The coevolution of society and multimedia technology: Issues in predicting the future innovation and use of a ubiquitous technology. Social Science Computer Review, 16(3), 268-282.

[10] Bajwa, D. S., Lewis, L. F., Pervan, G., Lai, V., Munkvold, B. E., & Schwabe, G. (2007). Organizational assimilation of collaborative information technologies: Global comparisons. In 40th Annual Hawaii International Conference on System Sciences (p. 41). Los Alamitos, CA: IEEE Computer Society

[11] Evers, Hans-Dieter (2008): Knowledge hubs and knowledge clusters: Designing a knowledge architecture for development. (ZEF Working paper Series, 27). Retrieved November 9, 2011, from www.zef.de/fileadmin/webfiles/downloads/zef_wp/wp27.pdf.

[12] Cummings, J. (January 5, 2007). Chris Dede on emerging technologies and neomillennial learning styles [Blog post]. Retrieved November 9, 2011, from www.educause.edu/blog/jcummings/ChrisDedeonEmergingTechnologie/166539

[13] Andriessen, D. (2004). Making sense of intellectual capital. Newton, MA: Butterworth-Heinemann. 

[14] Davenport, T.H., & Prusak, L. (2000). Working knowledge: How organizations manage what they know. Ubiquity, 1(24), 2.

[15] Frand, J., & Hixon, C. (December 1999). Personal knowledge management: Who, what, why, where, when, and how [Working paper]. Retrieved November 9, 2011, from www.anderson.ucla.edu/faculty/jason.frand/researcher/speeches/PKM.htm

[16] Pauleen, D. (2009). Personal knowledge management: Putting the "person" back into the knowledge equation. Online Information Review, 33(2), 221-224. Retrieved November 9, 2011, from www.ingentaconnect.com/content/mcb/264/2009/00000033/00000002/art00001

[17] Grundspenkis, J. (2007). Agent based approach for organization and personal knowledge modelling: Knowledge management perspective. Journal of Intelligent Manufacturing, 18(4), 451-7. Retrieved November 9, 2011, from www.springerlink.com/content/013277081u424052/

[18] Wright, K. (2005). Personal knowledge management: Supporting individual knowledge worker performance. Knowledge Management Research and Practice, 3(3), 156-65.


Denise Bedford is currently the Goodyear Professor of knowledge management at Kent State University. She is a member of several professional associations including ASIS&T, ACM, AIIM, SLA, DAMA, SCIP, ALA and AAAI. She can be reached at dbedfor3<at>kent.edu.