A decomposition cleanse function takes data from one field, breaks up the data into smaller pieces, and assigns the pieces to different columns in the staging table.  

For example, you have a salutation field that you want to break up into five separate name-type fields. The decomposition cleanse function has one input string and five output strings. In the mapping, you map the input string to the cleanse function, and you map each output string to the target columns in the staging table.

An aggregation cleanse function takes data from multiple fields, aggregates the data, and assigns the aggregated data to a single column in the staging table.

For example, you have five name-type fields that you want to aggregate into a single salutation field. The aggregation cleanse function has five input strings and one output string. In the mapping, you map the input strings to the cleanse function, and you map the output string to the target column in the staging table.