org.apache.ctakes.temporal.duration

Class Utils

java.lang.Object

org.apache.ctakes.temporal.duration.Utils

public class Utils
extends Object

Various useful classes and methods for evaluating event duration data.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	Utils.Callback Read event duration distributions from file.

Field Summary

Fields

Modifier and Type	Field and Description
static String[]	bins
static String	durationDistributionPath

Constructor Summary

Constructors

Constructor and Description
Utils()

Method Summary

Modifier and Type	Method and Description
static Map<String,Float>	convertToDistribution(String timeUnit) Take a time unit and return a probability distribution in which p(this time unit) = 1 and all others are zero.
static float	expectedDuration(Map<String,Float> distribution) Compute expected duration in seconds.
static String	formatDistribution(String mentionText, com.google.common.collect.Multiset<String> durationDistribution, String separator, boolean normalize) Convert duration distribution multiset to a format that's easy to parse automatically.
static org.apache.uima.collection.CollectionReader	getCollectionReader(List<File> inputFiles) Instantiate an XMI collection reader.
static List<File>	getFilesFor(List<Integer> patientSets, File inputDirectory) Get files for specific sets of patients.
static String	getPosTag(org.apache.uima.jcas.JCas systemView, org.apache.uima.jcas.tcas.Annotation annotation) Return system generated POS tag or null if none available.
static String	getTextBetweenAnnotations(org.apache.uima.jcas.JCas jCas, org.apache.uima.jcas.tcas.Annotation arg1, org.apache.uima.jcas.tcas.Annotation arg2) Get relation context.
static HashSet<String>	getTimeUnits(String timex) Extract time unit(s) from a temporal expression and put in one of the eight bins above.
static String	getTimexMLValue(String timex) Use Bethard normalizer to get TimeML value.
static String	getTimexMLValue(String timex, String anchorStr) Use Bethard normalizer to get TimeML value.
static String	lemmatize(String word, String pos) Lemmatize word using ClearNLP lemmatizer.
static void	main(String[] args)
static String	normalizeEventText(org.apache.uima.jcas.JCas jCas, org.apache.uima.jcas.tcas.Annotation annotation) Keep UMLS concepts and non-verbs intact.
static String	putInBin(String timeUnit) Take the time unit from Bethard noramlizer and return a coarser time unit, i.e.
static scala.collection.immutable.Set<java.time.temporal.TemporalUnit>	runTimexParser(String timex) Use Bethard normalizer to map a temporal expression to a time unit.
static void	writeInstance(String label, List<org.cleartk.ml.Feature> features, String fileName) Output label and list of cleartk features to a file for debugging.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

durationDistributionPath

public static final String durationDistributionPath

See Also:

Constant Field Values

bins

public static final String[] bins

Constructor Detail

Utils

public Utils()

Method Detail

getTimeUnits

public static HashSet<String> getTimeUnits(String timex)

Extract time unit(s) from a temporal expression and put in one of the eight bins above. Return empty set if time units could not be extracted. E.g. July 5, 1984 -> day

runTimexParser

public static scala.collection.immutable.Set<java.time.temporal.TemporalUnit> runTimexParser(String timex)

Use Bethard normalizer to map a temporal expression to a time unit.

getTimexMLValue

public static String getTimexMLValue(String timex)

Use Bethard normalizer to get TimeML value.

getTimexMLValue

public static String getTimexMLValue(String timex,
                                     String anchorStr)

Use Bethard normalizer to get TimeML value.

putInBin

public static String putInBin(String timeUnit)

Take the time unit from Bethard noramlizer and return a coarser time unit, i.e. one of the eight bins. Return null, if this cannot be done.

expectedDuration

public static float expectedDuration(Map<String,Float> distribution)

Compute expected duration in seconds. Normalize by number of seconds in a decade.

convertToDistribution

public static Map<String,Float> convertToDistribution(String timeUnit)

Take a time unit and return a probability distribution in which p(this time unit) = 1 and all others are zero. Assume time unit is one of the eight duration bins.

formatDistribution

public static String formatDistribution(String mentionText,
                                        com.google.common.collect.Multiset<String> durationDistribution,
                                        String separator,
                                        boolean normalize)

Convert duration distribution multiset to a format that's easy to parse automatically. Format: , :, ... Example: apnea, second:5, minute:1, hour:5, day:10, week:1, month:0, year:0

getTextBetweenAnnotations

public static String getTextBetweenAnnotations(org.apache.uima.jcas.JCas jCas,
                                               org.apache.uima.jcas.tcas.Annotation arg1,
                                               org.apache.uima.jcas.tcas.Annotation arg2)

Get relation context.

lemmatize

public static String lemmatize(String word,
                               String pos)
                        throws IOException

Lemmatize word using ClearNLP lemmatizer.

Throws:

IOException

getPosTag

public static String getPosTag(org.apache.uima.jcas.JCas systemView,
                               org.apache.uima.jcas.tcas.Annotation annotation)

Return system generated POS tag or null if none available.

normalizeEventText

public static String normalizeEventText(org.apache.uima.jcas.JCas jCas,
                                        org.apache.uima.jcas.tcas.Annotation annotation)
                                 throws org.apache.uima.analysis_engine.AnalysisEngineProcessException

Keep UMLS concepts and non-verbs intact. Lemmatize verbs. Lowercase before returning.

Throws:

org.apache.uima.analysis_engine.AnalysisEngineProcessException

getCollectionReader

public static org.apache.uima.collection.CollectionReader getCollectionReader(List<File> inputFiles)
                                                                       throws Exception

Instantiate an XMI collection reader.

Throws:

Exception

getFilesFor

public static List<File> getFilesFor(List<Integer> patientSets,
                                     File inputDirectory)

Get files for specific sets of patients. Useful for selecting e.g. only training files.

writeInstance

public static void writeInstance(String label,
                                 List<org.cleartk.ml.Feature> features,
                                 String fileName)

Output label and list of cleartk features to a file for debugging.

main

public static void main(String[] args)