"... In PAGE 7: ... Knowledge Content The knowledge content of the system was based on the requirements identified in the pre-implementation sur- vey. It consists of five major components, which correspond to different classifications of knowledge on the tacit/explicit dimension and the internal/external dimension (whether the knowledge was produced inside or out- side the organisation), as shown in Table1 . Unlike most knowledge management systems, which primarily focus on internal knowledge (Sveiby, 1997; Davenport et al, 1998; Davenport and Prusak, 1998; Hansen et al, 1999), this system is mainly focused on providing access to external knowledge, and in particular, the results of medical research.... ..."

"... In PAGE 6: ...here student.name=teacher.name and name= quot;John quot; and student.name in (select stu name from studies where course name = quot;A405 quot;) Example using DL Supposing that there exists a knowledge base (see Table1 ) with the classes course, person (with attribute name) and two subclasses teacher (with attributes teaches, teaches to and title) and student (with attribute studies). All the previous classes are primitive classes because they are described in terms of necessary conditions.... In PAGE 10: ... Most Immediate Superclasses. is fsuper teaching assistant,lucky teacher,atleast(15,studies)g, which veri es the previous properties: 1) super teaching assistant and lucky teacher belong to the in- tegrated view (see Table1 ) and atleast(15,studies) is a restriction of the query; 2) none of the elements of the set MIS subsumes another one and 3) the classes in the integrated view and elements in the query that subsume the class description are person, student, teacher, teaching assistant, super student, su- per teaching assistant, lucky teacher, atleast(15,studies) and atmost(0,teaches to). The rst ve elements subsume super teaching assistant.... In PAGE 12: ... A query q [rf(r) for] getall C is implicitly cached, denoted by q 2 ICQ, if all the classes whose names (Di) appear in the set MIS corresponding to the query q are explicitly or implicitly cached (getall Di 2 ECQ or getall Di 2 ICQ) and all the attributes that appear in the set MIS and the attributes to be projected in the query (rj) are also cached, that is, if rf(rj) for getall D 2 ECQ [ ICQ where D subsumes C. Therefore, if the sets Q and ECQ are known, the set ICQ is de ned in a recursive way as follows: ICQ = f q [rf(r) for] getall C j (q 2 Q) ^ (q = 2 ECQ) ^ (classes(MIS(q)) ECQ [ ICQ) ^ (8rl 2 (attributes(MIS(q)) [ frg) 9D ((rf(rl) for getall D 2 ECQ [ ICQ) ^ (D subsumes C)))g For example, using the knowledge base of Table1 , if the queries getall teacher, getall student and rf(studies) for getall student are explicitly cached then, the queries getall teaching assistant, getall super teaching assistant and getall su-... In PAGE 25: ... Version 1 evaluates all the possible states, Version 2 eliminates the states with more cost and less bene t and Version 3 eliminates the states with more cost and less maximum bene t. We have executed these versions with di erent sets of queries (formulated over the knowledge base of Table1 ) and parameter values, and we have seen that, on the one hand, Version 2 of the algorithm is the fastest one because it eliminates more states and Version 3 is the slowest one (see Table 3). This happens because the veri cation needed to eliminate a state in Version 3 is expensive10 and states are seldom eliminated.... ..."

"... In PAGE 22: ... The type of the object-level object is in turn characterised by the value of the referent attribute, which can be, for example, component (task, model or method) or project. Table2 lists the possible con gurations. In the case the object is a component, its componential type must also be speci ed.... ..."

"... In PAGE 6: ... These various technology systems described in Table1 form the collection of key data and information and then through various interactions of members of the multidisciplinary team with these technologies, protocols and treatment patterns are changed or developed. Table3 identifies each relevant case element in terms of the Knowledge Management Infrastructure Model presented earlier. What can be seen then, is a very heavy investment in the business intelligence infrastructure; i.... In PAGE 6: ... The Spine Unit has highly trained specialists who are encouraged to always keep at the cutting edge of new techniques for achieving better results and higher quality outcomes, with a strong emphasis on continuous improvement, they impart and exchange the knowledge and skills gained via interacting with the GDSS and the HIS components of the HMIS . From Table3 one can see that in this cure setting the knowledge management infrastructure is established and sustained through the technologies in place. By explicitly identifying the components of the knowledge management infrastructure, the specific knowledge processes, the supported tasks, the primary beneficiaries, and the specific examples from the Spine Unit case study, it is possible to make explicit the knowledge assets currently in place and thereby, facilitate better management of these knowledge assets, ensure better use and re-use of knowledge as well as maintain and update the knowledge management infrastructure itself.... ..."

"... In PAGE 9: ... The responses were favorable. The last question was meant to elicit the type of knowledge exchanged in the game rooms ( Table3 ). Various broad categories were pro- vided while also asking employees to state any explicit examples.... In PAGE 9: ...nd performance of the dot.com subsidiary. The big- gest section of the discussion concerned project teams. It is also interesting to note that all respondents who felt that they exchanged or received job-related infor- mation in the game rooms, said that they exchanged project team-related information ( Table3 ). Much of the discussions were to company-specific projects that were either in danger of being failures or running over budget.... ..."

"... In PAGE 14: ... The expressive power of the WebML E-R model can be compared to the WSML-Flight language. Table2 shows the comparison of the expressive power of WebML in respect to OWL-DL and WSML- Flight extending the comparison presented in [13]. Similar to WSML-Flight, the WebML data model comprises a rule language called WebML-OQL, which allows for calculating derived information, but not for defining constraints.... ..."

"... In PAGE 19: ... The data of the main regression variables are shown in Annex 1. Table1 shows that of the 111 countries in our sample, 69 countries (or 64 percent) have explicit deposit insurance. To our knowledge, this represents the complete list of countries with explicit deposit insurance as of year-end 2000.... ..."

"... In PAGE 9: ...chaotic, incohesive individuals. Table3 shows various forms of knowledge sharing and collabo- ration between concurrent teams, computers, and processes as discussed in [29]. Table 1: Forms of Sharing and Collaboration (after [29]).... In PAGE 16: ...Table3 presents explicit dependencies that can exist between agents, resources and tasks. Each cell represents a type of dependence that is a directed relation.... ..."

"... In PAGE 9: ...chaotic, incohesive individuals. Table3 shows various forms of knowledge sharing and collabo- ration between concurrent teams, computers, and processes as discussed in #5B29#5D. Table 1: Forms of Sharing and Collaboration #28after #5B29#5D#29.... In PAGE 16: ...Table3 presents explicit dependencies that can exist between agents, resources and tasks. Each cell represents atype of dependence that is a directed relation.... ..."

"... In PAGE 9: ...chaotic, incohesive individuals. Table3 shows various forms of knowledge sharing and collabo- ration between concurrent teams, computers, and processes as discussed in #5B29#5D. Table 1: Forms of Sharing and Collaboration #28after #5B29#5D#29.... In PAGE 16: ...Table3 presents explicit dependencies that can exist between agents, resources and tasks. Each cell represents atype of dependence that is a directed relation.... ..."