### Table 1. Graph search performance

in ABSTRACT

"... In PAGE 6: ...raph node. We used several grids of increasing size. For each grid, we performed a sequence of several searches, alternatively specifying two different viewing direction vectors. Table1 gives adjacency computation and search times as well as the number of sorted nodes per second as a function of grid size, for the CPU-GPU and software-based methods.... ..."

### Table 3: The effect of beam width pruning on word graph size and its accuracy; WGD is a measurement of word graph size

### Table 3: Measurements of width of star diagrams, in number of nodes at each level of the graph.

in Supporting the Restructuring of Data Abstractions through Manipulation of a Program Visualization

1998

"... In PAGE 25: ... Consequently, the middle of the diagram can be elided under user control to bring all the needed information onto the screen without loss of important details. Because a star diagram approaches its maximum width in just a few levels, major reductions in the size of a star diagram will only occur when pruned to the second or third level ( Table3 ). However, if the diagram can be interactively customized [10, 17], then temporarily eliding the diagram to such a severe level need not compromise understanding the diagram as a whole.... ..."

Cited by 34

### Table 6. Game graph, final approximation

"... In PAGE 9: ... Inserting these edges into the graph, we reach env and finally four new start states 000, 001, 100, and 101, and we obtain new summary edges to store in our tables. Continuing like this, we obtain the final approximation shown in Table6 and the summary edges for fA given in Table 11. We continue with an investigation on the main routine M given in Table 7.... ..."

### Table 2: Comparison over GraphBase directed graphs. All solutions 5 min.

2004

"... In PAGE 10: ...18 1.08 Table2 : Results for the rst method. 1 2 3 Q2 4 Q1 Q3 Q4 1 0 5 4 8 9 10 11 7 6 2 3 15 14 13 12 Figure 11: Literal nodes of 4-queens.... In PAGE 24: ... Problem class: BIBDs modelled as binary matrices. GAP-SBDD GAP-LEX Double Lex GAP-Lex no prop Combined V B R K 3 3 3 3 3 7 7 3 3 1 3 470 3 1150 3 20 21 1389 3 1150 6 10 5 3 2 4 869 29 80100 5 30 29 80100 4 50340 7 14 6 3 2 13 502625 - - 30 110 - - - - 9 12 4 3 1 12 451012 - - 30 120 - - - - 11 11 5 5 2 11 68910 - - 20 140 - - - - 8 14 7 4 3 14 219945 - - 143 720 - - - - - gt; 2 hours 3 Number of Backtracks Total runtime in ms Table2 : Table comparing various symmetry breaking methods. Partial GAP-LEX is where GAP-LEX checks do not commence until after the 1st backtrack.... In PAGE 24: ... Partial GAP-LEX is where GAP-LEX checks do not commence until after the 1st backtrack.GAP-SBDS GAP-SBDD Instance 3 2 4 3 2 4 K3 P2 9 290 110 400 22 310 180 490 K4 P2 165 1140 3590 4730 496 3449 8670 12110 K5 P2 4390 35520 166149 201669 17977 174180 501580 675760 GAP-LEX Partial GAP-LEX Instance 3 2 4 3 2 4 K3 P2 10 160 100 260 12 150 130 280 K4 P2 184 1550 4020 5570 202 670 4980 5650 K5 P2 4722 47870 176200 224070 5024 18820 224310 243130 3 Number of Backtracks 2 Gap Time in ms 4 Eclipse time in ms Total runtime in ms The results for this class of problems ( Table2 ) are more encouraging. We see that, in contrast to BIBDs, GAPLex provides fewer backtracks but performs faster than GAP-SBDD.... In PAGE 33: ...5 47.3 Table2 : Resolution times of the graph coloring problem. Figure 6: Convergencespeed to the best solution for the graph coloring problem with n = 30.... In PAGE 47: ... None Full Sibl. None Restarted no no no yes yes yes Table2 : Overview of the different algorithm variants: Full refers to the variant where we call for ancestor and sibling- based ltering at every search node. Sibling refers to breaking value symmetry only by performing just sibling-based lter- ing.... In PAGE 47: ... In case of the restarted method, the branching variable is chosen according to a min-domain heuristic over a random subset of 20% of the variables. Table2 summarizes the settings and names the different contestants that we let compete against one another. All experiments in this paper were conducted on a 2 GHz AMD Athlon 64 Processor 3000+ CPU with 512 MByte main memory running Linux 2.... In PAGE 62: ... We present in Table 1. and Table2 . results for var- ious graceful graph problems.... In PAGE 63: ...4 0 1986 139.4 Table2 : Results for computing one solution for graceful graphs Graph No sym break SBDS dynamic lex BT sec. BT sec.... ..."

### Table 2 Clustering coeSOcients of the market graph

2004

### Table 2 Comparison of the GIST andGMST on three diVTerent graphs a

"... In PAGE 14: ... A simple example will serve to illustrate this point. Table2 compares the weight andtortuosity of trees generated on several underlying graphs. The FFrst data set is for the graph used in generating Fig.... In PAGE 15: ...rom a Gaussian distribution with mean 0.5. The Gaussian distribution was scaled such that the standard deviation matched that of a uniform distribution between 0 and 1, with values below 0 or above 1 set to the appropriate extreme. The statistical properties of the GIST andGMST on this graph are listedin Table2 as Graph 3. The results are very similar to those for a graph with uniformly distributed edge weights.... ..."

### Table 1. Comparison over GraphBase directed graphs.

2005

"... In PAGE 4: ... The first set comes from [8]. The graphs are characterized by their probability (eta = 0:01 is noted r001 in Table1 ) that an edge is present between two distinct node n and n0. Those graphs were used to evaluate vflib algorithm performance [7].... In PAGE 4: ... Experiments show that CSP approach for subgraph matching solves more problem within a time limit against C++ specialized checking-based methods [7]. Table1 and 2 show the percentage of instances solved within a time limit of 5 minutes, for directed and undirected instances. Single specialized propagator MCPA for forbidden edges is more efficient than the version with two propagators.... ..."

Cited by 5