**1. Introduction**

184 New Research on Knowledge Management Technology

Umeki, H. (2007). A Challenge for computing in the 21st century: radwaste knowledge

19 April 2007, American Nuclear Society, LaGrange Park, IL (CD-ROM) Umeki, H.; Osawa, H.; Naito, M.; Nakano, K.; Makino, H. & McKinley, I.G. (2008).

Umeki, H.; Hioki, K.; Takase, H. & McKinley, I.G. (2009). Overview of the JAEA knowledge

Umeki, H. & Takase, H. (2010). Knowledge-based management systems for geological

Wenger, E.; McDermott, R. & Snyder, W. M. (2002). *Cultivating Communities of Practice*, ISBN-978-1-57851-330-7, Harvard Business Press, Boston, Massachusetts Yanev, Y. (Ed.) (2008). Special issue on nuclear knowledge management: IAEA perspective. *International Journal of Nuclear Knowledge Management*, Vol. 3, No. 2, 115–209

Liverpool, UK, 11–15 October 2009 (CD-ROM)

2007

109-2

2009 (CD-ROM)

*Conference on Environmental and Radioactive Waste Management*, ICEM2009-16355,

management, *Joint International Topical Meeting on Mathematics & Computation and Supercomputing in Nuclear Applications (M&C + SNA 2007)*, Monterey, CA, USA, 15–

Knowledge management: the cornerstone of a 21st century safety case, *OECD-NEA/EC/IAEA International Symposium on Safety Cases for the Deep Disposal of Radioactive Waste: Where Do We Stand?*, Paris La Défense, France, 23–25 January

management system supporting implication and regulation of geological disposal in Japan, *Proceedings of the ICEM'09: 12th International Conference on Environmental and Radioactive Waste Management*, ICEM2009-16354, Liverpool, UK, 11–15 October

disposal, In: *Geological Repositories for Safe Disposal of Spent Nuclear Fuels and Radioactive Materials*, Ahn, J. and Apted, M.J. (Eds.) Crc Pr I Llc, ISBN-978-1-43983Nowadays, it is commonly accepted that the World-Wide Web is the most popular tool for sharing knowledge and information (Berners-Lee, 1996). However, there is a huge and growing amount of information and it is getting more and more difficult to make sense out of it. The research work presented in this paper is an effort to address this Overload Information problem (Gross, 1964).

The main proposal of this work consists in managing the knowledge of a user community by means of a mechanism for knowledge construction in a distributed and incremental way. More specifically, knowledge evolves towards a structured and refined state by means of user interactions.

The aim of this mechanism is to achieve the crystallisation of user community knowledge as a result of user interactions, and without the need of an editor or manager of this task. The crystallised knowledge is the most accepted one by the community and, on the other hand, the knowledge that has not obtained enough acceptation will be likely to be removed.

It is important to highlight that knowledge is constantly evolving. Even crystallised knowledge may receive interactions from the community for further improvement. The key point is the evolution and improvement of knowledge by means of user evaluation.

The user community knowledge is subject to a maturation process involving two main phases. At first, due to the lack of critical mass of knowledge and interaction, a steering committee needs to be in charge of knowledge evaluation. Once enough mass is reached, knowledge crystallisation turns to be based on the evaluation performed by virtual communities of experts.

Those users who have added knowledge that has been crystallised are considered as experts, that is, their work have been recognised by the rest of the community. Virtual communities of experts are constructed in terms of sub areas of knowledge community, and they are in charge of the collaborative evaluation of the knowledge of their sub areas. This is similar to the peer review mechanism.

A collaborative Knowledge Management system called KnowCat has been designed and implemented ("Knowledge Catalyser"). KnowCat is based on the concept of Knowledge Crystallisation, supported by virtual communities of experts. KnowCat allows a user community to share, evaluate and structure collective knowledge. The system allows

Knowledge Crystallisation Supported by the KnowCat System 187

to be considered as the "best" description of the topic. This competitive environment is achieved by the Knowledge Crystallisation mechanism of the system, which is supported by virtual communities of users. Furthermore, each document has a "crystallisation degree", which determinates the social acceptation of this one for the user community (more details in Section 3). At any time, the author of a document can contribute with a new version of

Thirdly, each document can receive annotations –or note, for short–. A note is a review about the information presented in a document. Each note has a type that determinates its

a. "Clarification" note: this is useful to clarify some parts of the document. E.g. "The following link, that it appears in my document, it doesn't work now, but it worked a

b. "Support" note: this is useful to express agreement with the document. E.g. "This

c. "Review" note: this is useful to make suggestions about adding, removing, or changing some parts of the document, or for making comments regarding it. More specifically,

i. "Addition" note: to suggest additions to the document. E.g. "In my opinion, it is necessary to add in this document an index with its most important sections". ii. "Delete" note: to suggest deletions from the document. E.g. "In the summary there are

iii. "Correction" note: to suggest changes to the document. E.g. "I think that there is an error in the first paragraph of the conclusion section, it appears 'motor' instead of

iv. "Criticism" note: to criticise the document. E.g. "…Moreover the arguments are not

v. "Question" note: to make open questions about the document. E.g. "I think that the document author didn't express clearly his opinion about the document topic, please,

Finally, each document can receive assessments. An assessment represents a "weight assertion" which can be used by the users in order to determinate how good (with a value from 1, minimum value, to 10, maximum value) a specific aspect (i.e. correctness, innovation, etc.) of a specific part of a document (i.e. introduction, references, etc.) is. E.g. References.Completeness=9 *(Part.Aspect=value)* means that a specific document, in the opinion of the user, has the 90% of appropriated references (i.e. some few number of

**2.2 The collaborative work supported by KnowCat at the community knowledge** 

The KnowCat users can collaborate in a knowledge site trough the following potential interactions: modifying the knowledge tree, adding a document to a selected topic, voting a document, annotating and contributing with assessments about a document, adding a document version, accessing to documents and document versions, accessing to notes and

In figure 2, we can see an example screenshot of the community knowledge workspace of

document is very useful in my opinion and it is easy to read it".

some examples which, may be, are not necessary".

properly in our opinion in order to justify author position".

can you give us your opinion in the next document version?".

his/her document.

week ago ...".

'motivation'".

**workspace** 

assessments.

purpose. We have the following note types:

we have the following note types:

references are missing in this document).

"Technical Office" KnowCat site.

building Web sites where relevant and structured knowledge about some area or topic can be found (Cobos, 2003).

The KnowCat system is presented in the Section 2. Its Knowledge Crystallisation mechanism is detailed in Section 3. The system has evolved for the last twelve years, during this period it has been used with several user communities and a great amount of research data and results have been obtained, which are presented in Section 4. Finally, this paper concludes with some conclusions and future works in Section 5.
