Figure 4
A simple word count example to explain processing under the MapReduce framework. (a) The input is a series of words in a single document. (b) Each map scans part of the input. (c) An intermediate count of words is emitted by each map as a key-value pair. The word itself is the key and the count of its occurrence is the value. (d) Shuffle combines the intermediate key-values and passes them to reduce, where each reduce function obtains a distinct key along with all the intermediate values associated with that key. (e) Reduce computes the final count. |