AmilcareApi
Class API
java.lang.Object
|
+--AmilcareApi.API
- public class API
- extends java.lang.Object
Title: API for Amilcare
Description: Adaptive system for information extraction from text
Copyright: Copyright (c) 2000
Company: Dept. Computer Science Univ. of Sheffield
|
Constructor Summary |
API()
|
|
Method Summary |
static void |
extractInformation()
used to run Amilcare's rules on the current test corpus |
static void |
extractInformation(boolean showAnimation)
|
static java.lang.String |
getAmilcarePath()
it return Amilcare's operating directory |
static int |
getAnnotationAccuracy(java.lang.String tagName)
it returns the F-measure (a reasoned average of precision and recall) obtained for the annotation tagName (F is a number 0
|
static int |
getAnnotationActualMatches(java.lang.String tagName)
it returns the number of matches (either correct, missing or partial) for the annotation tagName |
static int |
getAnnotationCorrectMatches(java.lang.String tagName)
it returns the number of correct matches for the annotation tagName |
static java.lang.String[] |
getAnnotationList()
it returns an array of Strings that are the tags used by Amilcare for the current scenario |
static int |
getAnnotationMissingMatches(java.lang.String tagName)
it returns the number of missing matches for the annotation tagName |
static int |
getAnnotationPartialMatches(java.lang.String tagName)
it returns the number of partial matches for the annotation tagName. |
static int |
getAnnotationPossibleMatches(java.lang.String tagName)
it returns the number of instances present in the corpus for the annotation tagName |
static int |
getAnnotationPrecision(java.lang.String tagName)
it returns the precision obtained for the annotation tagName (prec is a number 0
|
static int |
getAnnotationRecall(java.lang.String tagName)
it returns the recall obtained for the annotation tagName (rec is a number 0
|
static int |
getAnnotationWrongMatches(java.lang.String tagName)
it returns the number of wrong matches for the annotation tagName |
static boolean |
getGazEnabled()
checks if the gazetteer is currently enabled |
static boolean |
getNercEnabled()
checks if the namedEntityRecognizer is currently enabled |
static boolean |
getPosEnabled()
checks if the pos tagger is currently enabled |
static boolean |
getPreEnabled()
checks if the preprocessor is currently enabled |
static java.lang.String |
getPreprocessedCorpusName()
The connection between Gate (preprocessing) and Amilcare is done through a file. |
static java.lang.String |
getRuleFileDir()
it returns the directory in which the rule file are stored by Amilcare |
static java.lang.String |
getScenarioFile()
returns the scenario file currently loaded |
static java.lang.String |
getScenarioName()
it returns the current scenartio mnemonic name |
static boolean |
getSplitCorpusInTwo()
|
static int |
getStatus()
it returns the status Amilcare is in. |
static AmilcareApi.IEResults[][] |
getTaggingResults()
It returns the tagging results as generated by Amilcare for external use. |
static AmilcareApi.IEResults[][] |
getTemplateResults()
It returns the Template results as generated by Amilcare for external use. |
static java.lang.String |
getTestCorpusName()
gets the current Test corpus name |
static java.lang.String |
getTrainingCorpusName()
gets the current training corpus name |
static void |
init()
|
static void |
learnRules()
used to learn rules from the current training corpus given the current scenario |
static void |
learnRules(boolean showAnimation)
|
static void |
loadScenario(java.lang.String fileName)
it reads a scenario file and sets Amilcare to (learn how to) extract information for that scenario
A scenario name is composed of a set of lines whose order is important:
1. |
static void |
main(java.lang.String[] args)
it tests the whole API |
static void |
preprocessCorpus()
it preprocesses the file in test or training coprus as defined by the scenario |
static void |
printTaggingResults()
it prints the tagging results in a formatted way |
static void |
printTemplateResults()
it prints the template results in a formatted way |
static void |
resetResults()
resets the system results so that in case of error the old results are not returned
as new. |
static void |
setAmilcarePath(java.lang.String value)
sets Amilcare's working directory. |
static void |
setAnnotationList(java.util.Collection annotList)
it sets the annotation list to be used by Amilcare. |
static void |
setDialogBoxesActive(boolean bool)
setDialogBoxesActive: tells Amilcare if the dialog boxes for errors are welcome (bool=true)
or unwelcome (i.e. |
static void |
setGazEnabled(boolean enable)
it allows to enable or disable the gazetteer in the preprocessing stage |
static void |
setNercEnabled(boolean enable)
it allows to enable or disable the NamedEntityRecognizer in Annie in the preprocessing stage |
static void |
setPosEnabled(boolean enable)
it allows to enable or disable the sentence splitter in the preprocessing stage |
static void |
setPreEnabled(boolean enable)
it allows to enable or disable the Gate-based preprocessing stage |
static void |
setPreprocessedCorpusName(java.lang.String fileName)
The connection between Gate (preprocessing) and Amilcare is done through a file. |
static void |
setRuleFileDir(java.lang.String dirName)
it sets the directory in which the rule file are stored by Amilcare |
static void |
setScenarioFile(java.lang.String filename)
sets the scenario file without loading it |
static void |
setScenarioName(java.lang.String name)
it sets the scenartio mnemonic name. |
static void |
setSplitCorpusInTwo(boolean value)
it asks amilcare to select half the corpus for learning and to use the other half for
testing. |
static void |
setSplitterEnabled(boolean enable)
it allows to enable or disable the sentence splitter in the preprocessing stage |
static void |
setTestCorpusName(java.lang.String fileName)
sets the current Test corpus file to the parameter |
static void |
setTrainingCorpusName(java.lang.String fileName)
sets the current training corpus file to the parameter |
static void |
stop ()
used to stop Amilcare when running. |
| Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
taggingResults
public static AmilcareApi.IEResults[][] taggingResults
templateResults
public static AmilcareApi.IEResults[][] templateResults
API
public API()
resetResults
public static void resetResults()
- resets the system results so that in case of error the old results are not returned
as new.
init
public static void init()
getTaggingResults
public static AmilcareApi.IEResults[][] getTaggingResults()
- It returns the tagging results as generated by Amilcare for external use.
- Returns:
- an array whose elements are texts in the corpus (if only one text was provided it returs a one-el array). Each text is represented by an array of
IEResults structures
Caveat! some of the elements can be null (both at text level and at IEResults structures)
printTaggingResults
public static void printTaggingResults()
- it prints the tagging results in a formatted way
getTemplateResults
public static AmilcareApi.IEResults[][] getTemplateResults()
- It returns the Template results as generated by Amilcare for external use.
- Returns:
- an array whose elements are texts in the corpus (if only one text was provided it returs a one-el array). Each text is represented by an array of
IEResults structures
Caveat! some of the elements can be null (both at text level and at IEResults structures)
printTemplateResults
public static void printTemplateResults()
- it prints the template results in a formatted way
setSplitterEnabled
public static void setSplitterEnabled(boolean enable)
throws java.lang.Exception
- it allows to enable or disable the sentence splitter in the preprocessing stage
- Throws:
java.lang.Exception - if the module is not available
setPosEnabled
public static void setPosEnabled(boolean enable)
throws java.lang.Exception
- it allows to enable or disable the sentence splitter in the preprocessing stage
- Throws:
java.lang.Exception - if the module was not available
getPosEnabled
public static boolean getPosEnabled()
- checks if the pos tagger is currently enabled
setGazEnabled
public static void setGazEnabled(boolean enable)
throws java.lang.Exception
- it allows to enable or disable the gazetteer in the preprocessing stage
- Throws:
java.lang.Exception - if the module is not available
getGazEnabled
public static boolean getGazEnabled()
- checks if the gazetteer is currently enabled
setNercEnabled
public static void setNercEnabled(boolean enable)
throws java.lang.Exception
- it allows to enable or disable the NamedEntityRecognizer in Annie in the preprocessing stage
- Throws:
java.lang.Exception - if the module is not available
getNercEnabled
public static boolean getNercEnabled()
- checks if the namedEntityRecognizer is currently enabled
setDialogBoxesActive
public static void setDialogBoxesActive(boolean bool)
- setDialogBoxesActive: tells Amilcare if the dialog boxes for errors are welcome (bool=true)
or unwelcome (i.e. they are disruptive wrt the external application strategy: bool=false). If unwelcome the errors and all the messages are
written on the std output. It is duty of the calling system to tell the user about the problem
setPreEnabled
public static void setPreEnabled(boolean enable)
throws java.lang.Exception
- it allows to enable or disable the Gate-based preprocessing stage
- Throws:
java.lang.Exception - if the module is not available
getPreEnabled
public static boolean getPreEnabled()
- checks if the preprocessor is currently enabled
getRuleFileDir
public static java.lang.String getRuleFileDir()
- it returns the directory in which the rule file are stored by Amilcare
setRuleFileDir
public static void setRuleFileDir(java.lang.String dirName)
- it sets the directory in which the rule file are stored by Amilcare
loadScenario
public static void loadScenario(java.lang.String fileName)
- it reads a scenario file and sets Amilcare to (learn how to) extract information for that scenario
A scenario name is composed of a set of lines whose order is important:
1. name of the scenario (e.g. seminarAnnouncement)
2. a list of tags (with angular brackets) (one for each line) introduced by the tag <__++SLOTS>
and ended by </__++SLOTS>. For example:
<__++SLOTS>
<speaker>
<stime>
<etime>
<location>
</__++SLOTS>
3. the type of task to be performed (e.g. Information Tagging and Template FIlling)
4. reserved for future use
5. the name of the training corpus (e.g. D:\Fabio\Java\Tagging\Interface\Data\Corpus\1seminar.xml)
6. reserved for future use
7. the name of the test corpus (e.g. D:\Fabio\Java\Tagging\Interface\Data\Corpus\1seminar.xml)
8. the number of texts to be analysed from the corpus (e.g.: 10)
9. the random seed to be used in ncross folder experiments (e.g. 0) (advanced use only -- put 0 if unsure)
10. true or false, according if the end of line is an interesting features to be used by Amilcare (advanced users only -- leave false if unsure)
11. the directory where Amilcare will put the induced rules (e.g. D:\Fabio\Java\Tagging\Interface\Data\Rules\)
12. reserved for future use
13. reserved for future use
14. reserved for future use
15. an int: the lenght of the window for rule pattern (e.g. 4). Please note that this is the window to the either left or right of the tag
if 4 is specified the rule will put 4 words to the left and four to the right (i.e. 8 words)!!
16. a double indicating the maximum error threshold for rules (e.g. 0.3333333333333333) (advance use only -- set 0.33333 if unsure)
17. a double for the minimum error threshold for rules (e.g. 0.03333333333333333)(advance use only -- set 0.033333 if unsure)
18. an int: the minimum number of matches for a tagging rule to be accepted (e.g. 1) (advance use: put 1 if unsure)
19. a double: the balance between precision and recall for tagging rules (advance use: put 1.0 if unsure)
20. an int: the minimum number of matches for a correction rule to be accepted (e.g. 1) (advance use: put 1 if unsure)
21. a double: the balance between precision and recall for correction rules (advance use: put 1.0 if unsure)
22. the name of the file where the preprocessed corpus is stored (as returned by getPreprocessedCorpus)
- Returns:
- nothing, as side effect it calls
setScenarioFile
getScenarioFile
public static java.lang.String getScenarioFile()
- returns the scenario file currently loaded
- Returns:
- a complete file path
setScenarioFile
public static void setScenarioFile(java.lang.String filename)
- sets the scenario file without loading it
getTrainingCorpusName
public static java.lang.String getTrainingCorpusName()
- gets the current training corpus name
setTrainingCorpusName
public static void setTrainingCorpusName(java.lang.String fileName)
- sets the current training corpus file to the parameter
getTestCorpusName
public static java.lang.String getTestCorpusName()
- gets the current Test corpus name
setTestCorpusName
public static void setTestCorpusName(java.lang.String fileName)
- sets the current Test corpus file to the parameter
getPreprocessedCorpusName
public static java.lang.String getPreprocessedCorpusName()
- The connection between Gate (preprocessing) and Amilcare is done through a file.
This file is rewritten automatically each time amilcare is run either for training or
for IE. If the file is given each time a different name, it can be reused in order to save
the GAte's processing time. This is particularly useful when the same document can be processed
more than once.
This method allows to retrieve the filename where Amilcare is writing the intermediate results
setPreprocessedCorpusName
public static void setPreprocessedCorpusName(java.lang.String fileName)
- The connection between Gate (preprocessing) and Amilcare is done through a file.
This file is rewritten automatically each time amilcare is run either for training or
for IE. If the file is given each time a different name, it can be reused in order to save
the GAte's processing time. This is particularly useful when the same document can be processed
more than once.
This method allows to set the filename where Amilcare is writing the intermediate results
the calling program is expected to check for filename validity!
getAnnotationList
public static java.lang.String[] getAnnotationList()
- it returns an array of Strings that are the tags used by Amilcare for the current scenario
setAnnotationList
public static void setAnnotationList(java.util.Collection annotList)
- it sets the annotation list to be used by Amilcare.
extractInformation
public static void extractInformation()
- used to run Amilcare's rules on the current test corpus
extractInformation
public static void extractInformation(boolean showAnimation)
preprocessCorpus
public static void preprocessCorpus()
- it preprocesses the file in test or training coprus as defined by the scenario
learnRules
public static void learnRules()
- used to learn rules from the current training corpus given the current scenario
learnRules
public static void learnRules(boolean showAnimation)
stopAmilcare
public static void stopAmilcare()
- used to stop Amilcare when running. Please take into account that stopping can take a while
getScenarioName
public static java.lang.String getScenarioName()
- it returns the current scenartio mnemonic name
setScenarioName
public static void setScenarioName(java.lang.String name)
- it sets the scenartio mnemonic name.
setSplitCorpusInTwo
public static void setSplitCorpusInTwo(boolean value)
- it asks amilcare to select half the corpus for learning and to use the other half for
testing. The selection is pseudo-random (currently Amilcare selects every other text
for learning and all the others for training--- this strategy is not to be taken for granted!!!)
obviously Amilcare still uses the maximum number of texts to be considered
(i.e. for learning and testing it considers texts from half the corpus up to the maximum number of texts)
getSplitCorpusInTwo
public static boolean getSplitCorpusInTwo()
getAnnotationCorrectMatches
public static int getAnnotationCorrectMatches(java.lang.String tagName)
- it returns the number of correct matches for the annotation tagName
getAnnotationMissingMatches
public static int getAnnotationMissingMatches(java.lang.String tagName)
- it returns the number of missing matches for the annotation tagName
getAnnotationPartialMatches
public static int getAnnotationPartialMatches(java.lang.String tagName)
- it returns the number of partial matches for the annotation tagName. A partial
match is defined here as a match for which either the start is correct or the end
(we do not consider partial something that overlaps in any other way)
getAnnotationWrongMatches
public static int getAnnotationWrongMatches(java.lang.String tagName)
- it returns the number of wrong matches for the annotation tagName
getAnnotationPossibleMatches
public static int getAnnotationPossibleMatches(java.lang.String tagName)
- it returns the number of instances present in the corpus for the annotation tagName
getAnnotationActualMatches
public static int getAnnotationActualMatches(java.lang.String tagName)
- it returns the number of matches (either correct, missing or partial) for the annotation tagName
getAnnotationPrecision
public static int getAnnotationPrecision(java.lang.String tagName)
- it returns the precision obtained for the annotation tagName (prec is a number 0
getAnnotationRecall
public static int getAnnotationRecall(java.lang.String tagName)
- it returns the recall obtained for the annotation tagName (rec is a number 0
getAnnotationAccuracy
public static int getAnnotationAccuracy(java.lang.String tagName)
- it returns the F-measure (a reasoned average of precision and recall) obtained for the annotation tagName (F is a number 0
setAmilcarePath
public static void setAmilcarePath(java.lang.String value)
- sets Amilcare's working directory. it is equivalent to setting the AmilcarePath
in the java command (java -d amilcarepath=value)
TO be effective it must be set AFTER Amilcare's init!
getAmilcarePath
public static java.lang.String getAmilcarePath()
- it return Amilcare's operating directory
getStatus
public static int getStatus()
- it returns the status Amilcare is in. It is used to produce a progress bar in the interface
main
public static void main(java.lang.String[] args)
- it tests the whole API