Table 1 Feature extraction subsystem
From: KLOSURE: Closing in on open–ended patient questionnaires with text mining
Module | Software | Resources | Output |
|---|---|---|---|
linguistic pre-processing | Stanford Core NLP [38] | language model | POS tags, dependencies |
sentiment analysis | Stanford Core NLP [40] | sentiment model | sentiment polarities |
named entity recognition | MetaMap [31] | UMLS [21] | named entities, semantic types |
lexicon lookup | N/A | 24 lexicons | matched items |