→ Back to personalknowledge.org
PKM2009 — First International Workshop on Personal Knowledge Management (PKM)
at the 5th Conference on Professional Knowledge Management,
25-27 March 2009, Solothurn, Switzerland
Organizers: Max Völkel, Heiko Haller, Dr. Andreas Abecker; all at Abt. Wissensmanagement FZI Forschungszentrum Informatik at the Karlsruhe Institute of Technology, Germany
Bookmark URL: http://2009.personalknowledge.org (this page)
Info page: http://semanticweb.org/wiki/PKM2009 (frozen)
The workshop had an acceptance rate of 50% for long papers and 80% for all papers.
ca. 30 participants
PKM2009 is with 35 registered participants third-largest workshop at the conference.
PKM2009 was the first academic workshop dedicated to Personal Knowlege Management (PKM).
Knowledge Management (KM) deals with creating and exchanging knowledge within groups of persons in organisational contexts. The potentials and needs of the individual is often not in the focus of KM efforts, although no-one would deny that the individual as knowledge bearer, -user and -creator is naturally the most essential part of knowledge management. The main goal of PKM is make the individual more productive - and thereby the organisation as a whole. The term Personal Knowledge Management (PKM) has two main dimensions:
- Personal Knowledge
Ultimately, all knowledge is personal knowledge. Following the tradition of Nonaka&Takeuchis spiral model (and later Ba model) knowledge resides partially in the minds of people and can partially be codified as external artifacts. PKM investigates the use of methods and tools to amplify the abilities of the individual to work better with knowledge. E.g.
recall previously learned knowledge faster (or at all) when it is required
model personal knowledge and beliefs with external modeling tools to derive new insights (MS Excel is often used for this today)
strategies for filing ideas to retrieve them when needed
- Personal Management
Management is a systematic approach to define goals, measure, define and execute actions and repeat this control loop until the goal is reached. Different from traditional management, in personal management one has to manage oneself. This involves the problem of fulfilling two roles (executing and managing) and learning when and how to switch between them. Typical management problems in PKM are e.g.
time and task management
matching work habits with personal productivity level variations
investing time into personal learning and PKM improvements
Max Völkel, David Elsweiler, Matthias Hert, Lars Ludwig, Carsten Waldeck, Heiko Haller, Benjamin Adrian, Marucs Liwicki, Andreas Abecker and Richard Pircher
Traditionally, Knowledge Management (KM) deals with creating and exchanging knowledge within groups, especially within organisational contexts. The needs and potential of individuals are rarely the focus of KM, although nobody could deny that individual knowledge workers, as creators, locators, posessors and consumers of knowledge, are the most essential element in KM. In Personal Knowledge Manamgent (PKM) the focus of attention is on individual knowledge workers and the aim of the research in this area is to increase the productivity of these individuals and, as a by-product, increasing the productivity of groups and organisations. Switching the focus to individuals raises a number of challenges and opens knowledge management problem to other domains interested in similar, perhaps overlapping, problems.
The aim of the two day workshop was to explore the challenges and issues of PKM and move towards the creation of a research community interested in working in this area. Papers submitted to the workshop were reviewed by an international programme committee and 5 full papers and 3 short papers were accepted for the workshop proceedings. These papers reflected the diversity of interest and challenges in the field of PKM, describing research on interface technology, infrastructure and modelling, evaluation, mapping between different knowledge structures, as well as cognitive and neurobiological perspectives on the problem.
The workshop began with presentations of the accepted papers. The first session featured full papers and the second session featured short-papers.
In his talk, "Towards Ontology-based Information Extraction and Annotation of Paper Documents for Personalized Knowledge Acquisition", Benjamin Adrian showed a method for bridging the media gap between physical documents and digital desktops. Along a semi-automatic annotation process, each document is annotated with Concepts of the Personal Information Model Ontology and finally transformed into a Semantic Wiki Article. Benjamin showed a user interface that visualizes an interactive document image, where users may accept recommended annotations or create new annotations manually. As base technology, the whole approach mainly grounds on Open Source solution i.e. Gnowsis as Semantic Desktop, Ocropus as OCR Service, Kaukolu Wiki as Semantic Wiki, and finally a free but not open source Information Extraction Engine iDocument.
David Elsweiler raised the issue of evaluation for Personal Information and Knowledge Management. He argued that this has been a major hurdle in the Personal Information Management community where he has been active. He presented his project which aims at triangulating laboratory and naturalistic evaluations using mathematical modelling techniques.
Marcus Liwicki presented an extension of the Semantic Desktop with a novel input modality: handwriting on paper. This extension allows for annotating printed documents with handwritten notes, which are then recognized, interpreted and integrated into the personal knowledge space. Thus it supports personal knowledge work on paper. In particular, the user can mark any text segment and annotate it with his or her own comments. The semantic of these comments and marks are then interpreted and, finally, the result is included in the user’s Semantic Desktop. One may argue that the trend is going towards the paperless office. However, currently this is more a myth than reality. The presented system can be seen as a supporting tool for the intermediate step we are currently in. Furthermore, it has to be noted that the system also supports work on electronic paper-devices.
In his presentation, Max Völkel emphasised the need to examine the costs and benefits of taking notes for personal knowledge management. He introduced the concept of a knowledge cue - an object or piece of information that has the potential to remind an individual about a particular piece of knowledge, but does not necessarily have to contain knowledge itself. He further stressed the need to let the user decide on the degree of granularity, formality and structured-ness of knowledge cues within a personal knowledge model.
Matthias Hert addressed the benefits Semantic Web technologies can have in Personal Information and Knowledge Management to exchange and integrate data from various sources across application and system boundaries. He presented an approach to overcome the heterogeneity problems that arise through the existence of multiple vocabularies for the same application domain. He introduced a mapping language and a transformation solution to bridge between these different data representations.
Olaf Grebner presented an architecture framework for higher-level activity support using multiple functional workspaces on the desktop. The presented architecture framework tackles the problem that today’s desktop systems face the two major drawbacks in supporting the knowledge worker (KWer) of a) insufficient support for a KWer’s higher level activities and b) the scattering of a KWer’s personal information across the desktop and its applications. However, research hasn’t yet shown an integrated architecture for realizing multiple workspaces on such a common abstract information model. The presented architecture framework proposes to base these functional-oriented workspaces on a common personal information model which itself is maintained as service. By leveraging the Nepomuk Social Semantic Desktop as foundational system and example for an underlying service-oriented desktop architecture, the framework represents an evolutionary architecture in contrast to more radical, but not commercially implemented research approaches.
The final two papers brought two very different perspectives on PKM. For example, Richard Pircher, in his talk, broadened the perspective, viewing meta-cognition and self-reflexion as central to PKM. By meta-cognition he meant the rational thinking about thought processes and knowledge about one’s knowledge and he meant self-reflexion in the sense of the individual’s more global dealing with his own identity. He also referred to recent neuropsychological research showing that emotions are so tightly interwoven with other cognitive activities, that they must be regarded even when dealing with 'objective' fact knowledge.
Lars Ludwig introduced 3 basic forms of a person’s artificial memory. 'Reflective artificial memory' directly reflecting and thus supporting consciousness. 'Extending artificial memory' reflecting what was once thought or experienced and thus - in principal - could also be remembered from biological memory. And 'expanding artificial memory' comprising new information that has to be associated to reflective or extending artificial memory in a meaningful way. Artificial memory systems are to be used as personal proxies that help reflecting, remembering, and learning. Lars Ludwig showed that knowledge can stem both from biological memory and from perception (of artificial memory). Both ways of knowing (thinking and remembering versus perceiving and being reminded) are prone to a great number of disruptions preventing from knowing. In (personal) knowledge management, not enough attention is being paid to these disruptions.
These papers proved to be a fertile source of discussion for the remainder of the workshop.
As a means to generate further discussion points, participants in the workshop were encouraged to demo research prototypes that they were working on. These were the system demos given
Lars Ludwig showed his tool "artificial memory" that runs in a web browser and can also be tried out online at http://www.artificialmemory.net . It acts as a finegrained personal knowledge database that can be browsed in a tree-like fashion. Concepts can be added and interlinked with semantic relations. The system also has imported several thesauri so that some semantic search can be done where e.g. the sub-class information of concepts is considered. While its usage may not be very intuitive at first glance, watching Lars using it shows that – once accustomed to it – it can be quite a powerful and expressive tool.
Heiko Haller demoed his graphical knowledge mapping tool "iMapping" and the built-in semantic search tool "QuiKey". The current iMapping application is a stand-alone desktop application that allows to build large knowledge maps with deeply nested items that can be zoomed into. While it can be used as an informal notetaking tool comparable to a large pinpoard, iMapping also allows to interlink items with semantic links to build more formal personal knowledge models. The meaning of these relations can be refined in QuiKey, something like a semantic command-line tool for more tech-savvy users, that allows not only fast incremental fulltext search but also to browse and bulid structured queries on the personal knowledge base. iMapping and QuiKey are downloadable and described in more detail at imapping.info.
Marcus Liwicki showed the current prototype of Semantic eInk, which is able to recognize annotations is a specified format. While the functionality is already quite large (printing on any documents, perfect mapping of gestures to text), the handwriting recognition is not ready for practical use. This is the focus on the ongoing work on this prototype.
Matthias Hert demonstrated the Web interface of RDFTransformer, a prototype implementation of a transformation service for RDF data. It enables the definition of mappings from a source ontology (e.g., vCard) to a target ontology (e.g., FOAF) in the RDF Data Transformation Language (RDTL) and converts RDF data based on such mappings. RDFTransformer is implemented as a library for embedding in other application but also provides simple command line and Web interfaces. Another component is a mapping server that can provide common mappings over a local network or the internet to multiple instances of the RDFTransformer. The project website http://seal.ifi.uzh.ch/RDFTransformer (dead link) provides more information about RDFTransformer, the Web interface, and the prototype for download.
The first session of the second day of the workshop included an invited talk by Carsten Waldeck from Liquiverse. Carsten introduced some of the novel information access interfaces his company has been designing and outlined how they may impact on personal knowledge management. The interfaces are based on the concept of liquid browsing, which aims at optimizing information gain per pixel. The user can easily browse data on a scatterplot and change the displayed dimensions - numeric and ordinal. Carsten believes that these kinds of interfaces will empower the user as they struggle with the task of managing their information and knowledge.
The last sessions of both days were dedicated to a group discussion of the issues raised within the workshop. A highly productive discussion ensued relating to a broad range of topics. However, there were four main threads to the discussion. We summarise these below:
- Paper vs. digital
Despite all efforts to manage life with computers, paper notes are used by allmost all people.
- Personal vs. group vs. organisation
There was some tension between the approach to PKM and the distinction between these categories. There were also differing viewpoints within the group of participants regarding whether focusing on the "personal" was just a useful way of examining organisational knowledge management or whether it is a completely separate problem and domain. Lars Ludwig contributed to the debate by introducing legal distinctions between the categories and underlining how changing work practises are and will continue to have an influence.
- Knowledge vs. Information vs. Data
Although the focus of the workshop was on the management of knowledge, the conversation often turned to information and sometimes to data. It was clear that there is often overlap between these concepts and that the management of knowledge must involve the management of information. However, a strong arguement in favour of focusing on knowledge lies with the value. Information is not something of particular value and can in fact be costly in terms of time, attention etc. Knowledge on the other hand, is extremely valuable and worth managing.
Evaluation was a theme discussed heavily throughout the workshop. David Elsweiler asked the provocative question "is innovation enough to be scientific?", refering to the current situation where many systems had been proposed, but very few evaluated. The difficulities involved in evaluation were debated including controlling variables, managing privacy, attracting participants and lack of suitable evaluation methods and metrics. The participants were in agreement that evaluation was important and started making progress about what was important to measure (example metrics) and ways that measurements could be established. It was clear that this is an area in need of further research.
- Personal Practice
Stimulated through the talks by Lars Ludwig and Richard Pircher, there were periods in the workshop when the discussion became very philosophical in nature.
One of the issues was the claim that practices like meditation can indirectly help PKM e.g. rising awareness about one’s own cognitive processes and patterns and by reducing cognitive over-activity by training the mind to relax and focus.
Throughout the workshop the discussion demonstrated that Personal Knowledge Management is both a complex area and one that requires input from many sources. It was evident throughout that this is a multi-disciplinary field and that related research is being performed in a wide range of academic areas. However, it was also clear from the discussion that there is a group of people interested in forming a community to study this problem. We are, therefore, grateful for the input from our presenters, attendees, and contributors to the proceedings who gave us two lively days of discussion and provided insights from different academic perspectives.
14.00 - 15.30: PKM2009, I
Introduction of Participants
Paper Presentations (10 minutes presentation per full paper)
- Towards Ontology-based Information Extraction and Annotation of Paper Documents for Personalized Knowledge Acquisition
Benjamin Adrian, Heiko Maus, Malte Kiesel and Andreas Dengel
- Evaluating Personal Information Management Behaviour using Markov Models
David Elsweiler, Stefan Mandl and Leif Azzopardi
- A Paper-based Technology for Personal Knowledge Management
Kinga Schumacher, Marcus Liwicki and Andreas Dengel
- Personal Knowledge Models for More Productive Knowledge Workers
15.30 - 16.00: Break
16.00 - 17.30: PKM2009, II
Paper Presentations (5 minutes presentation per short paper)
- Personal Knowledge Mapping with Semantic Web Technologies
Matthias Hert, Gerald Reif and Harald Gall (full paper, 10 minutes) (slides online)
- Architecture Framework for Higher-Level Activity Support Using Multiple Functional Workspaces on the Desktop
- Persönliches Wissensmanagement als Management der Metakognition
- Artificial Memory - Eine kurze Einführung in Struktur, Aufgaben und Erfolgskennzahlen
Position statements of participants without papers (ca. 2 minutes per participant)
Discussion and Informal Tool Demo Session - bring PKM tools You wrote or use!
17.30 - 18.00: Break
18.00 - 20.00: Conference Poster-Session and Get together
20.00: PKM workshop dinner at Zum Wirthen
08:30 - 09:30: Conference Keynote
09:30 - 11:00: PKM2009, III
- Invited Talk: "Liquid Browsing in Personal Data Spaces"
Carsten Waldeck, designer, inventor and entrepreneur (Liquiverse GmbH)
11:00 - 11:30: Break
11:30 - 13:00: PKM2009, IV
Presentation of results
PKM workshop lunch