"... In PAGE 7: ... Final results are the averages over these 10 runs. They are listed in Table2 . On one dataset, housing, all three methods are compatible.... In PAGE 7: ... All 3 learning methods are applied to the 7 datasets. Results for the average of 10 runs of random samples are listed in Table2 . In general, accuracies in Ta- ble 3 are slightly worse than those in Table 1, as expected.... In PAGE 7: ... We applied this dimension reduction (DR) ver- sion of CF and the results are shown in Table 4. The rst 2 columns are for the 10% labeled case as in Table2 . The second 2 columns are for the noise label case as in Table 3.... ..."

"... In PAGE 2: ... That is, both of these knowledge representations have reached a certain level of maturity, which make them suitable candidates for combination. Among frameworks for uniting rules and DLs (see Table1 ), homogeneous ap- proaches { like DLP (Grosof et al. 2003), SWRL (Horrocks et al.... In PAGE 2: ... 2007) { can be distinguished, cf. the double-bar in Table1 , from hybrid approaches, like AL-log (Donini et al. 1998), CARIN (Levy and Rousset 1998), HEX-programs (Eiter et al.... In PAGE 17: ...o be safe, i.e., a variable that appears in the head must also appear in the body { we call it as the Datalog safeness condition in this paper, and the above undecidable encoding is such a case. As mentioned in Table1 , CARIN, DLP and SWRL obey this Datalog safeness, but either CARIN or DLP has its respective restrictions under other considerations as to obtain decidability, while SWRL admits itself undecidable.... ..."

"... In PAGE 2: ... The framework, described in [19] and later enhanced in [18, 17], in particular addresses model comprehensibility, but also includes social quality and knowledge quality; conform to a social constructivistic approach to process support. Our framework, depicted in Figure 2 and elabo- rated in Table1 , is based on the following concepts: A business process is represented in a business process model expressed in a process modeling language (PML). The model is subject to audience interpretation from various human stakeholders and technical actors (i.... In PAGE 3: ... Model quality types were formalized in [17] based on viewing the model, the domain, the interpretation and the participant knowledge all as sets of statements. The defi- nition of the various model quality types with related goals, means and accompanying modeling activities are summarized in Table1 . Means that apply at several levels are only stated the first time, and language quality goals are looked upon as means in the framework.... In PAGE 6: ...1 Model quality in surveyed products Physical quality has two goals: externalization and internalizability; cf. Table1 . Externalization is closely linked to language quality, and is covered later.... ..."

"... In PAGE 2: ... The framework, described in [19] and later enhanced in [18, 17], in particular addresses model comprehensibility, but also includes social quality and knowledge quality; conform to a social constructivistic approach to process support. Our framework, depicted in Figure 2 and elabo- rated in Table1 , is based on the following concepts: A business process is represented in a business process model expressed in a process modeling language (PML). The model is subject to audience interpretation from various human stakeholders and technical actors (i.... In PAGE 3: ... Model quality types were formalized in [17] based on viewing the model, the domain, the interpretation and the participant knowledge all as sets of statements. The defi- nition of the various model quality types with related goals, means and accompanying modeling activities are summarized in Table1 . Means that apply at several levels are only stated the first time, and language quality goals are looked upon as means in the framework.... In PAGE 6: ...1 Model quality in surveyed products Physical quality has two goals: externalization and internalizability; cf. Table1 . Externalization is closely linked to language quality, and is covered later.... ..."

"... In PAGE 8: ...nd exemplified in subsection 5.2.2. Table1 lists the metrics that are taken into account, and presents how maturity levels are assigned. The general design of the analysis framework is presented in Figure 3.... In PAGE 9: ... However, the accountability and responsibility for the second activity are assigned just as stated in Cobit (50 % assigned, 50 % assigned according to Cobit). According to Table1 , this results in assignment to maturity levels 2 and 3 regarding the activity execution metric for the two activities respectively. The equivalent activity level maturity for the three activities is calculated according to Table 2.... In PAGE 10: ...organization. This equals 60 % or maturity level 3 according to Table1 . Nevertheless, just two of the suggested KPI:s/KGI:s are continuously monitored, which is equivalent of 22 % or maturity level 1.... ..."