
Similarity Join for Big Geographic Data 35
After generating the groups in a reduce node, the MapReduce
framework calls the reduce function Reduce_base once for each group. This
function is presented in Algorithm 5. The function receives as input the
key-value pair (k2, v2List). k2 is the intermediate key of one of the records
of the group being processed and v2List is the list of values of all the
records of the group. If the size of the list is small enough to be processed
in a single node, the algorithm calls a single-node Similarity Join routine,
i.e., InMemorySimJoin, to get the links in the current partition (lines 1 to 2).
Otherwise all ...