"... In PAGE 14: ... The optimal result would be fapp(g; g; g); app(g; nong; nong); app(nong; g; nong); app(nong; nong; nong)g, but our procedure will return the most general model having all eight possible combinations of g, nong arguments. We show in Table1 the abstract models of certain predicates extracted from the CiaoPP database. Table 2 shows the derived models over the FTA defln- ing the types dynamic and static (which denote the same as gnd and term in the standard models, but are the names used in the binding time analysis of Logen).... ..."

"... In PAGE 5: ... We conjecture that INT could serve as a powerful alternative proof engine for verifying the abstract model during abstraction refinement. Table2 shows the number of registers in the abstract models generated by the three abstraction refinement methods. The first column shows the names of the properties.... In PAGE 6: ... Apparently the scoring of each input of the abstract models and the construction of the SORs bogged down the performance of the method. Table2 also shows that between 33 and 100 percent of the variables that CNTL included in the abstract models are both k- controllable and k-cooperative. The rest of the variables are k- cooperative only.... ..."

"... In PAGE 7: ... Agent systems within a region have to register with the server, which maintains access control lists for region resources and monitors agent systems and agents in a centralized way. Table2 provides an overview of the mapping between our abstract model and Jumping Beans. 4.... ..."

"... In PAGE 7: ... Agent systems within a region have to register with the server, which maintains access control lists for region resources and monitors agent systems and agents in a centralized way. Table2 provides an overview of the mapping between our abstract model and Jumping Beans. 4.... ..."

"... In PAGE 7: ... Agent systems within a region have to register with the server, which maintains access control lists for region resources and monitors agent systems and agents in a centralized way. Table2 provides an overview of the mapping between our abstract model and Jumping Beans. 4.... ..."

"... In PAGE 7: ... Agent systems within a region have to register with the server, which maintains access control lists for region resources and monitors agent systems and agents in a centralized way. Table2 provides an overview of the mapping between our abstract model and Jumping Beans. 4.... ..."

"... In PAGE 7: ... Agent systems within a region have to register with the server, which maintains access control lists for region resources and monitors agent systems and agents in a centralized way. Table2 provides an overview of the mapping between our abstract model and Jumping Beans. 4.... ..."

"... In PAGE 5: ...The first phase of any usage scenario consists of generating an abstract model corresponding to the Famix (FAMOOS6 Information Exchange Model) specifi- cation. The contents of this model (see Table2 ) requires input from (i) Source Navigator; (ii) CCCC; and (iii) home-made script to extract namespace scopes and conditional compilation direc- tives. The facts provided by these three tools are combined and interconnected by SNAVTOFAMIX7, resulting in the gener- ation of a CDIF (Case Data Interchange Format [Che03]) representation of the Famix model.... ..."

"... In PAGE 9: ... PE1 and PE2 are micro-processors while PE3 and PE4 are custom-hardwares. Table1 summaries the characteristics of different ab- straction models.... ..."