The mapper is the first step in MapReduce framework, a component of a larger scalable, parallel-izable algorithm.
Maps input key/value pairs to a set of intermediate key/value pairs.
Maps are the individual tasks which transform input records into a intermediate records. The transformed intermediate records need not be of the same type as the input records. A given input pair may map to zero or many output pairs.
The most common map reduce framework is Apache Hadoop.
See also MapReduce Wiki.