### Table 1: Accuracy of the tree edit distance and its approximations.

2005

"... In PAGE 10: ... It is computed in the same way as the pq-gram distance, the only di erence being that the pro le of a tree consists of the bag of all its node labels. The results for the address tables RO and LR are shown in Table1 . There are two streets in RO that do not exist in LR, thus jMcj = 300 for the calculation of the accuracy.... ..."

Cited by 2

### Table 2: Longest Common Subsequence of the First Product

### Table 1: Time for approximate matching in seconds. m is the pattern length and k is the edit distance.

2001

"... In PAGE 5: ...We used a Sun Ultra 450 workstation (400MHz CPU, 4G bytes memory). Table1 shows the time for approximate matching with error level 5%.... ..."

Cited by 14

### Table 1: Edit distance calculation (arranged by CUPS/chip).

1993

"... In PAGE 4: ... However, the e ort in programming these systems is signi cantly lower than designing a VLSI chip, and they are general-purpose, able to perform a large number of functions. Table1 summarizes the performance of several of these machines. The number of PEs, maximum native sequence length, and the approximate number of chips are listed for each machine.... In PAGE 5: ... Jones reports 75 MCUPS on a 64K CM-2 (corresponding to two times faster than the results in the table) by microcoding the inner loop of the dynamic programming algorithm [19]; Jones has also presented methods for database pattern searching with limited gap length on the CM-2 [18]. By making full use of modulo sequence comparison to reduce data communication, this author estimates that a factor of 2{4 performance increase is attainable over the CM-2 and MP-1 results of Table1 for long sequences [16]. Core has compared dynamic programming and the BLAST algorithm on the CM-1 and Intel iPSC hypercube computer [8].... ..."

Cited by 6

### Table 1: Edit distance calculation (arranged by CUPS/chip).

1993

"... In PAGE 4: ... However, the e ort in programming these systems is signi cantly lower than designing a VLSI chip, and they are general-purpose, able to perform a large number of functions. Table1 summarizes the performance of several of these machines. The number of PEs, maximum native sequence length, and the approximate number of chips are listed for each machine.... In PAGE 5: ... Jones reports 75 MCUPS on a 64K CM-2 (corresponding to two times faster than the results in the table) by microcoding the inner loop of the dynamic programming algorithm [19]; Jones has also presented methods for database pattern searching with limited gap length on the CM-2 [18]. By making full use of modulo sequence comparison to reduce data communication, this author estimates that a factor of 2{4 performance increase is attainable over the CM-2 and MP-1 results of Table1 for long sequences [16]. Core has compared dynamic programming and the BLAST algorithm on the CM-1 and Intel iPSC hypercube computer [8].... ..."

Cited by 6

### Table 1: Time for approximate matching in seconds. m is the pattern length and k is the edit distance.

"... In PAGE 5: ...We used a Sun Ultra 450 workstation (400MHz CPU, 4G bytes memory). Table1 shows the time for approximate matching with error level 5%.... ..."

### Table 1: Time for approximate matching in seconds. m is the pattern length and k is the edit distance.

"... In PAGE 5: ...We used a Sun Ultra 450 workstation (400MHz CPU, 4G bytes memory). Table1 shows the time for approximate matching with error level 5%.... ..."

### Table 1: True similarity between sequences and the length of longest common subsequence found by using K randomly chosen linear functions; averages over 10 trials. Data: two series of 85 points about the Finnish national economy.

1997

"... In PAGE 3: ... Moreover, it produced approximations to the true similarity that are very close to the correct val- ues. For example, in Table1 we see that for varying , the randomized algorithm got to within 1 from the true optimum already after 500 randomly chosen linear func- tions. Table 2 shows the time needed for this analysis.... ..."

Cited by 25

### Table 1: True similarity between sequences and the length of longest common subsequence found by using K randomly chosen linear functions; averages over 10 trials. Data: two series of 85 points about the Finnish national economy.

1997

"... In PAGE 10: ... Moreover, it produced approximations to the true similarity that are very close to the correct values. For example, in Table1 we see that for varying , the randomized algorithm got to within 1 from the true optimum already after 500 randomly chosen wedges. Table 2 shows the time needed for this analysis.... ..."

Cited by 25

### Table 2: Cost of tree editing operations for the abstraction (Costa) and refinement (Costr) approximation scores

2007

"... In PAGE 16: ...pproximation score (i.e., their costs have to be equal to 0). Table2 illustrates the costs of the tree edit distance operations for abstraction and refinement approximation scores. This extension influences the tree edit distance computation.... ..."