Skip to main content

Table 7 Best parameters for optimizing performance for precision or recall

From: Large-scale biomedical concept recognition: an evaluation of current automatic annotators and their parameters

High precision annotations

NCBO Annotator

 

MetaMap

 

ConceptMapper

 

Parameter

Value

Parameter

Value

Parameter

Value

wholeWordOnly

YES

model

STRICT

searchStrategy

CONTIGUOUS

filterNumber

ANY

gaps

NONE

caseMatch

SENSITIVE

stopWords

ANY

wordOrder

ORDER MATTERS

stemmer

NONE

SWCaseSensitive

ANY

acronymAbb

DEFAULT/UNIQUE

stopWords

NONE

minTermSize

THREE/FIVE

derivationalVariants

NONE

orderIndLookup

OFF

withSynonyms

NO

scoreFilter

1000

findAllMatches

NO

  

minTermSize

3/5

synonyms

EXACT ONLY

High recall annotations

NCBO Annotator

 

MetaMap

 

ConceptMapper

 

Parameter

Value

Parameter

Value

Parameter

Value

wholeWordOnly

NO

model

RELAXED

searchStrategy

SKIP ANY/ALLOW

filterNumber

ANY

gaps

ALLOW

caseMatch

IGNORE/INSENSITIVE

stopWords

ANY

wordOrder

IGNORE

stemmer

Porter/BioLemmatizer

SWCaseSensitive

ANY

acronymAbb

ALL

stopWords

PubMed

minTermSize

ONE/THREE

derivationalVariants

ALL/ADJ NOUN

orderIndLookup

ON

withSynonyms

YES

scoreFilter

0

findAllMatches

YES

  

minTermSize

1/3

synonyms

ALL