### Table 3. Comparison of retrieval performance for CACM.

2000

"... In PAGE 18: ... To evaluate the retrieval effectiveness of the resulting ranked document lists, we used eight different evaluation measures. The results, averaged over the set of queries, are displayed in Table3 and Table 4. If we look at the differences in performance between any two of the three methods, we get the following rough indications.... ..."

Cited by 12

### Table 5. Most cited Journals in ER papers up to 1999. Journal Title # Cites ACM Transactions on Database Systems 886 Communications of the ACM 278

"... In PAGE 4: ... Source names in Table 4 follow the journal abbreviations used by the Web of Science. See Table5 for a list of the most popular ones. Table 4 is predominated by journal papers, except the two ER papers by Scheuermann and Santos.... In PAGE 4: ... Table 4 is predominated by journal papers, except the two ER papers by Scheuermann and Santos. Table5 lists most frequently cited journals in ER papers between 1979 and 1999. Table 1.... ..."

### Table 3 Communication Characteristics of Parallel LT Algorithms Algorithm Messages Data Volume

1997

"... In PAGE 18: ... FFT/Trans. LT: PY lt; 2 3(J + 1) dK=PXe K=PX 1 + 9 4 P 2 Y (J + 1)2 ! dK=PXe K=PX 1 + 9 4 P 2 Y (J + 1)2 ! PY 23(J + 1) dK=PXe K=PX 6PY 2J + 5 dK=PXe K=PX 6PY 2J + 5 Adapting (17) and (18) to this situation and using 2J 3M + 1, we obtain the expres- sions in Table3 for LT-related communication costs when using transpose algorithms. The transpose LT algorithms, like the transpose FFT algorithms, become less e - cient when modi ed to overlap computation with communication.... ..."

Cited by 30

### Table 3: Communication Characteristics of Parallel LT Algorithms Algorithm Messages Data Volume

"... In PAGE 23: ... Hence, assuming perfect load balance, D = 16(M +1)JK=PX for the transpose preceding the forward LT and D = 10(M +1)JK=PX for the transposition following the inverse LT. Adapting (18) and (19) to this situation and using 2J 3M + 1, we obtain the expressions in Table3 for LT-related communication costs when using transpose algorithms. The transpose LT algorithms, like the transpose FFT algorithms, become less e cient when modi ed to overlap computation with communication.... ..."

### Table 1 Reduction in communication volume using Binsort

"... In PAGE 5: ... This algorithm carries out approximate sorting of the grid points in O(N) time, and does not require communication between processors, provided that the excessive amount of work Ep is known to each processor p. As shown in Table1 , it has been observed that the communication volume can be reduced by up to 53 percent by applying the Binsort algorithm. It should be noted that the e ect of the Binsort algorithm on the number of messages transmitted is negligible, because it does not change the relationship between donor and donee processors.... ..."

### Table 1 Reduction in communication volume using Binsort

in Chapter 1 Runtime Support and Dynamic Load Balancing Strategies for Structured Adaptive Applications

"... In PAGE 5: ... This algorithm carries out approximate sorting of the grid points in O(N) time, and does not require communication between processors, provided that the excessive amount of work Ep is known to each processor p. As shown in Table1 , it has been observed that the communication volume can be reduced by up to 53 percent by applying the Binsort algorithm. It should be noted that the e ect of the Binsort algorithm on the number of messages transmitted is negligible, because it does not change the relationship between donor and donee processors.... ..."

### Table 1: communication complexity pattern round volume

"... In PAGE 6: ... 3.6 Efficiency Table1 shows communication pattern, round complexity, and data volume per round of each step of our protocol. Here, l is the number of links, m is the number of nodes, v is the maximum number of links for a node, and n is the number of components of an encrypted weight that must be greater than the length of the longest path.... ..."

### Table 6: Volume of communication (2 PEs).

1997

Cited by 2

### Table 2: Communication and Disk Access Volume for the Outside Communication pattern

"... In PAGE 6: ...s shown in Section 4.1. We did the analysis for the other five disk placement orders as above. The results are shown in Table2 and details can be obtained from [11]. 4.... In PAGE 6: ... We did the analysis for the other five disk placement orders as above. The results are shown in Table2 and details can be obtained from [11]. 5 Overhead Minimization for the Inside Com- munication Method In this section, we analyze each of the three parallel algo- rithms possible with the inside communication pattern and determine the minimal Overhead cost achievable along with the conditions under which this will be possible.... ..."

### Table 5 Coefficients from the regression of volume of communication (phone and phone plus email) on wired status and other independent variables (N=42).

2003

"... In PAGE 18: ... Netville residents with access to the local computer network made 4 to 5 times as many phone calls to social ties of all strengths. [Table 4] The regression model showing the volume of telephone communication for those residents talked with on a regular basis is similar to the regression model for ties recognized by name ( Table5 ). Length of residence and being wired are associated with the volume of phone communication.... ..."

Cited by 25