A Node Like Any Other
“Nodes” are what SAS calls their text analytics sub-tasks of their Text Miner® application for deriving business intelligence from unstructured text. There are several nodes, of which the following are some major ones:
There is the Import Node which routes text for possible anaysis into a storage area where the actual analytics takes place. The Topic Node is where the information of highest potential value is kept, and this is where summary reports of the analytic results are produced.
The Text Parsing node takes any text that is analyzed as having little value to the particular project and stages it for elimination. Also performed in this node is automated spell checking and down-stemming (the structural decomposition of compound words into its root). And finally, the Filter Node takes outputs from the parsing node and breaks down the results into various metrics such as frequency counts.
For a full description of SAS Text Miner analytics software, see the full product review.