|
MetaMap Transfer (MMTx) |
Rosetta Stone
Usage:
java programs.Tokenize [Options]
The Tokenizer tokenizes collections into documents,
documents into sections, and finally sections into sets of
Sentences and Tokens.
This tokenizer recognizes MEDLINE Citation documents and
tokenizes sections according to the structure of MEDLINE citations,
input formated for the SMART retrieval system, fielded text,
and free text.
The options include the following:
Input and Ouput File Options
Short Name
Long Name
Default Value
Purpose
__
--fileName=
stdIn
Name of file to process
__
--outputFileName=
stdOut
Name of the outputFile to write to
Input Format Descriptions
The default behavior is to auto-detect MEDLINE Citation format or free
text. The following flags overwrite this feature.
Short Name
Long Name
Default Value
Purpose
__
--medlineCitations
false
The input is a collection of medLine citations
__
--mrcon
false
The input is a collection of MRCON rows
__
--freeText
true
The input is free text
__
--fieldedText
false
Is the input file/stdin fielded text?
__
--textField=
2
For fielded text, which field contains the text
__
--fieldSeparator=
|
For fielded text, what char is the separator
Options to retrieve ever more levels of detail
Short Name
Long Name
Default Value
Purpose
__
--collections
false
Display Collection information
__
--documents
false
Display Documents
__
--sections
false
Display Sections
__
--sentences
false
Display Sentences
__
--tokens
false
Display tokens
__
--pipedOutput
false
Display in a pipe delimited format
__
--details
false
Display the goory details
Processing Options
Short Name
Long Name
Default Value
Purpose
__
--ambiguousAcronyms
false
Disambiguate sentence boundries using the acronyms and
abbreviations file
Configuration Options
Short Name
Long Name
Default Value
Purpose
__
--configName=
mmtx cfg
The name of the configuration file
-R
--MMTX_ROOT=
/export/nls/mmtx
MMTX Root path
__
--ambiguousAcronymsFile=
data/lexicon/ambiguousAcronymsFile.txt
Location of the acronyms and abbreviations file needed in
the tokenizer.
__
--nmm
false
Flag that flips between MetaMap output and non MetaMap output
sytle. This flag is useful when combined with the --pipedOutput
and display flags such as the --sentences, --phrases, --nps,
--variants and other levels of detail.
