### Table I. Complete list of the selected atom types, with their PDB codes

2003

Cited by 9

### Table 3: Merge matrix for an atomic object.

"... In PAGE 6: ... The particular policy of Table 2 is to request user selection if one user modified an element and the other deleted i~ insertions of either user are always accepted. Merge matrices are also defined for simple types such as integers and character stings ( Table3 ). Such merge matrices... ..."

### Table 1 and the top level qualities Adaptability, Security and Availability. These are respectively marked ++, +, ++ for the selected style. More fine-grain qualities have been identified during the decomposition process, such as Integrity (Accuracy, Completeness), Usability, Response Time, Maintainability, and more.

2001

"... In PAGE 6: ... Such, admittedly complex, contracting arrangements can be used to manage conditions of complexity and uncertainty deployed in high- cost-high-gain (high-risk) applications. s-i-5 pyr jo-ve bidd ar-le hi-co co-op Predictability + ++ + -- - - Security + ++ + -- -- - Adaptability + ++ ++ + + ++ Cooperativity + ++ + - - + ++ Competitivity - - - ++ ++ + + Availability + + ++ - -- + -- Failability-Tolerance -- -- ++ - Modularity ++ - + ++ + + -- Aggregability ++ ++ + Table1 . : Correlation catalogue.... In PAGE 6: ... For lack of space, we do not detail them here and refer to [12] where a full description of such attributes is presented. Table1 summarizes correlations for our styles and the quality attributes: +, ++, -, -- respectively model partial/positive, sufficient/positive, partial/negative and sufficient/negative contributions [3]. 3 Multi-agent patterns A further step in the architectural design of MAS consists of specifying how the goals delegated to each actor are to be fulfilled.... ..."

Cited by 62

### Table 4. Sampling Strategies for Selecting Informants (Adapted from Patton, 2002)

2004

"... In PAGE 17: ... Importantly, in selecting interviewees, case researchers must ensure that the sampling strategy is consistent with the purpose of the inquiry. In order to help IS researchers make informed decisions, we present below some of the most common informants sampling strategies (see Table4 ) suggested by Patton (2002). First, maximum variation sampling occurs when an investigator seeks to obtain the broadest range of information and perspectives on the subject of study.... ..."

Cited by 2

### Table 1. Atomic operation definitions (adapted from [2]).

"... In PAGE 7: ... We are defining a complete set of generic authoring tasks corresponding to all three information y-layers (subject domain, course, and library), which are supported by the domain engine, course engine and library engine respectively. These tasks are defined on the base of a set of atomic operations, which are abbreviated and summarised in Table1 . Due to lack of space, we cannot present in this paper the full ontology of courseware authoring tasks we developed so far.... ..."

### Table 6. Decomposition_obj Matrix

2003

"... In PAGE 9: ... quot; quot; ,$ amp; apos; + and % $, ) # $ amp; apos; + (Table 2, 3, 4, and 5). An other example is to create a traceability matrix between instances of the and classes, called ! quot; #$# % amp; apos; ( ( Table6 ), of the Requirement Management sub-model (Figure 1). The SR model describes the intentional relationships that are internal to actors, such as means-ends relationship and task decomposition [8].... In PAGE 9: ... The SR model describes the intentional relationships that are internal to actors, such as means-ends relationship and task decomposition [8]. The components that participated in the means-ends relationship and task decomposition are mapped as follows: the tasks and softgoals elements in the ! quot; #$# % amp; , - matrix (Table 3) and the tasks and goals elements in the ! quot; #$# % amp; apos; ( matrix ( Table6 ). For example, Table 2 indicates that the placement of an order (RF03) requires the use (realized by) of a Shopping Cart (RF13).... In PAGE 9: ...nvolves (i.e. it is AND decomposed into) the selection of item (RF14), its addition to the cart (RF15), proceeding to check it out (RF16) and providing the relevant identification details (RF17). In turn, Table6 , indicates that there are two means (OR - alternatives) of getting the identification information (RF17), either using the internet (SO06) or using classic means such as sending a fax (SO05). The matrixes are also useful to find which functional requirements are related some non-functional requirement.... ..."

Cited by 2

### Table 1 The addition of o axis basis functions in the planes perpendicular to the internuclear axis and passing through the atomic centres to an axis basis set containing only atom-centred basis functions. The di erences between the calculated and exact energy values are given in parentheses.

"... In PAGE 8: ...Table1 consists of a 30s subset centred on each of the atoms supplemented by 28s subsets, obtained by deleting the two most di use functions from the 30s set, centred on the eight o axis atom centre, oa(ac), positions de ned by (8) with n1 = 5: 3 Results and Discussion The results of a series of calculations in which o axis basis subsets were added to an axis basis set consisting of only atom-centred basis function are summa- rized in Table 1. In this Table EmHF denotes the matrix Hartree-Fock energy obtained with the speci ed basis set.... In PAGE 8: ...28s subsets, obtained by deleting the two most di use functions from the 30s set, centred on the eight o axis atom centre, oa(ac), positions de ned by (8) with n1 = 5: 3 Results and Discussion The results of a series of calculations in which o axis basis subsets were added to an axis basis set consisting of only atom-centred basis function are summa- rized in Table1 . In this Table EmHF denotes the matrix Hartree-Fock energy obtained with the speci ed basis set.... In PAGE 9: ...The addition of o axis basis functions in the plane perpendicular to the inter- nuclear axis and passing through the bond centre to an axis basis set consisting of both atom-centred and bond-centred basis functions is explored in Table 2. Comparing the errors associated with the 30s ac; 30s bc basis set with those of the 30s ac set presented in Table1 , it can be seen that the error in the matrix Hartree-Fock energy is reduced by 5013:3 Hartree to only 1:5% of the error measured for the 30s ac set whereas for the electron correlation energy the error is reduced by only 1474:6 Hartree to 90:4% of the reference value in Table 1. This illustrates the e ectiveness of bond centred basis subsets in reducing the errors matrix Hartree-Fock energies but the relative ine ectiveness of such sub- sets in signi cantly reducing the errors in calculated correlation energies.... In PAGE 9: ... The ma- trix Hartree-Fock energy is reduced by 65 Hartree on adding four o axis basis subsets in the plane perpendicular to the internuclear axis and passing through the bond centre to the 30s ac; 30s bc basis set. Thus, for n0 1 = 5, the error in the matrix Hartree-Fock energy is 12:0% of that in the reference energy in Table 2 and 0:2% of the error in the reference energy in Table1 . For the second order correlation energy the error is reduced by 9799:3 Hartree with respect to the reference energy in Table 2 when n0 1 = 4: The corresponds to 29:5% of the error in the reference in Table 2 and 26:6% of the error in the reference in Table 1.... In PAGE 13: ... A 20s basis set centred on each nucleus yields a matrix Hartree-Fock energy of ?1:1284986 Hartree and a sec- ond order correlation energy of ?18:496 mHartree. Thus the errors in the ma- trix Hartree-Fock energy and the second order energy are comparable with that recorded in Table1 for the 30s ac basis set. A 20s10p basis set lowers the matrix Hartree-Fock energy by 5:026 mHartree to ?1:1335251 Hartree and the second order correlation energy by 13:252 mHartree to ?31:748 mHartree.... ..."

### Table 3: E ciency of the representation. Problem Working basis Factorization Update

"... In PAGE 20: ...exchange. Table3 reports CPU time in seconds of the factorization and the average time of a single update for representative working bases in the earlier selected linear programs. The results reported here were obtained by running our FORTRAN implementation of the method on a 33MHz SUN SPARC 10 workstation.... In PAGE 20: ... Additionally, in the factorization time we distinguish time spent in the SPK1 heuristic to reorder the matrix to a bordered block triangular form, torder, time spent to compute the Schur matrix (6), tS, time to decompose S to the explicit LU factors, tLU and their sum tFactor = torder + tS + tLU. Analysis of Table3 leads to the following observations. The factorization step is quite e cient unless the Schur matrix grows to a large size in which case the computation of S and its LU decomposition become dominating terms.... ..."

### Table 3 The addition of o axis basis functions in the planes perpendicular to the internuclear axis and passing through the atomic centres and the bond centre to an axis basis set consisting of both atom-centred and bond-centred basis functions. The di erences between the calculated and exact energy values are given in parentheses.

"... In PAGE 10: ...133 620 710 (8: 86) -0.029 264 511 (4613:22) The addition of o axis basis functions in the planes perpendicular to the in- ternuclear axis and passing through the atomic centres and the bond centre to an axis basis set consisting of both atom-centred and bond-centred basis functions is considered in Table3 . The reference energies with respect to which the results presented in Table 3 should be assessed correspond to the 30s ac; 30s bc basis set which is also used as a reference in Table 2.... In PAGE 10: ...029 264 511 (4613:22) The addition of o axis basis functions in the planes perpendicular to the in- ternuclear axis and passing through the atomic centres and the bond centre to an axis basis set consisting of both atom-centred and bond-centred basis functions is considered in Table 3. The reference energies with respect to which the results presented in Table3 should be assessed correspond to the 30s ac; 30s bc basis set which is also used as a reference in Table 2. The addition of eight sets of o axis functions in planes passing through the atomic centres is not as e ective as the four sets of o axis functions in the plane passing through the bond centre in reducing the error in the matrix Hartree-Fock energy.... ..."

### Table 4 The addition of secondary o axis basis functions in the planes perpendicular to the internuclear axis and passing through the atomic centres and the bond centre to an axis basis set consisting of both atom-centred and bond-centred basis functions. The di erences between the calculated and exact energy values are given in parentheses.

"... In PAGE 11: ...133 621 535 (8:035) -0.032 451 958 (1425:773) The e ects of adding secondary o axis basis functions are summarized in Table4 . The results obtained with the largest basis set considered in Table 3, 30s ac; 30s bc; 28s oa(ac) [n1 = 4] 26s oa(bc) [n0 1 = 4], are used as a reference with respect to which the results presented in Table 4 are analyzed.... In PAGE 11: ...032 451 958 (1425:773) The e ects of adding secondary o axis basis functions are summarized in Table 4. The results obtained with the largest basis set considered in Table 3, 30s ac; 30s bc; 28s oa(ac) [n1 = 4] 26s oa(bc) [n0 1 = 4], are used as a reference with respect to which the results presented in Table4 are analyzed. The addition of secondary o axis basis functions reduces the error in the matrix Hartree-Fock energy to 4 Hartree.... ..."