Appendix F
Sample ML Configuration File [#]
<?xml version="1.0" encoding="UTF-8"?>
<ML-CONFIG> <DATASET> <!-- The type of annotation used as instance --> <INSTANCE-TYPE>Token</INSTANCE-TYPE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>Lookup(0)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Lookup</TYPE> <!-- The position relative to the instance annotation --> <POSITION>0</POSITION> </ATTRIBUTE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>Lookup_MT(-1)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Lookup</TYPE> <!-- Optional: the feature name for the feature used to extract values for the attribute --> <FEATURE>majorType</FEATURE> <!-- The position relative to the instance annotation --> <POSITION>-1</POSITION> <!-- The list of permitted values. if present, marks a nominal attribute; if absent, the attribute is numeric (double) --> <VALUES> <!-- One permitted value --> <VALUE>address</VALUE> <VALUE>cdg</VALUE> <VALUE>country_adj</VALUE> <VALUE>currency_unit</VALUE> <VALUE>date</VALUE> <VALUE>date_key</VALUE> <VALUE>date_unit</VALUE> <VALUE>facility</VALUE> <VALUE>facility_key</VALUE> <VALUE>facility_key_ext</VALUE> <VALUE>govern_key</VALUE> <VALUE>greeting</VALUE> <VALUE>ident_key</VALUE> <VALUE>jobtitle</VALUE> <VALUE>loc_general_key</VALUE> <VALUE>loc_key</VALUE> <VALUE>location</VALUE> <VALUE>number</VALUE> <VALUE>org_base</VALUE> <VALUE>org_ending</VALUE> <VALUE>org_key</VALUE> <VALUE>org_pre</VALUE> <VALUE>organization</VALUE> <VALUE>organization_noun</VALUE> <VALUE>person_ending</VALUE> <VALUE>person_first</VALUE> <VALUE>person_full</VALUE> <VALUE>phone_prefix</VALUE> <VALUE>sport</VALUE> <VALUE>spur</VALUE> <VALUE>spur_ident</VALUE> <VALUE>stop</VALUE> <VALUE>surname</VALUE> <VALUE>time</VALUE> <VALUE>time_modifier</VALUE> <VALUE>time_unit</VALUE> <VALUE>title</VALUE> <VALUE>year</VALUE> </VALUES> <!-- Optional: if present marks the attribute used as CLASS Only one attribute can be marked as class --> </ATTRIBUTE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>Lookup_MT(0)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Lookup</TYPE> <!-- Optional: the feature name for the feature used to extract values for the attribute --> <FEATURE>majorType</FEATURE> <!-- The position relative to the instance annotation --> <POSITION>0</POSITION> <!-- The list of permitted values. if present, marks a nominal attribute; if absent, the attribute is numeric (double) --> <VALUES> <!-- One permitted value --> <VALUE>address</VALUE> <VALUE>cdg</VALUE> <VALUE>country_adj</VALUE> <VALUE>currency_unit</VALUE> <VALUE>date</VALUE> <VALUE>date_key</VALUE> <VALUE>date_unit</VALUE> <VALUE>facility</VALUE> <VALUE>facility_key</VALUE> <VALUE>facility_key_ext</VALUE> <VALUE>govern_key</VALUE> <VALUE>greeting</VALUE> <VALUE>ident_key</VALUE> <VALUE>jobtitle</VALUE> <VALUE>loc_general_key</VALUE> <VALUE>loc_key</VALUE> <VALUE>location</VALUE> <VALUE>number</VALUE> <VALUE>org_base</VALUE> <VALUE>org_ending</VALUE> <VALUE>org_key</VALUE> <VALUE>org_pre</VALUE> <VALUE>organization</VALUE> <VALUE>organization_noun</VALUE> <VALUE>person_ending</VALUE> <VALUE>person_first</VALUE> <VALUE>person_full</VALUE> <VALUE>phone_prefix</VALUE> <VALUE>sport</VALUE> <VALUE>spur</VALUE> <VALUE>spur_ident</VALUE> <VALUE>stop</VALUE> <VALUE>surname</VALUE> <VALUE>time</VALUE> <VALUE>time_modifier</VALUE> <VALUE>time_unit</VALUE> <VALUE>title</VALUE> <VALUE>year</VALUE> </VALUES> <!-- Optional: if present marks the attribute used as CLASS Only one attribute can be marked as class --> </ATTRIBUTE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>Lookup_MT(1)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Lookup</TYPE> <!-- Optional: the feature name for the feature used to extract values for the attribute --> <FEATURE>majorType</FEATURE> <!-- The position relative to the instance annotation --> <POSITION>1</POSITION> <!-- The list of permitted values. if present, marks a nominal attribute; if absent, the attribute is numeric (double) --> <VALUES> <!-- One permitted value --> <VALUE>address</VALUE> <VALUE>cdg</VALUE> <VALUE>country_adj</VALUE> <VALUE>currency_unit</VALUE> <VALUE>date</VALUE> <VALUE>date_key</VALUE> <VALUE>date_unit</VALUE> <VALUE>facility</VALUE> <VALUE>facility_key</VALUE> <VALUE>facility_key_ext</VALUE> <VALUE>govern_key</VALUE> <VALUE>greeting</VALUE> <VALUE>ident_key</VALUE> <VALUE>jobtitle</VALUE> <VALUE>loc_general_key</VALUE> <VALUE>loc_key</VALUE> <VALUE>location</VALUE> <VALUE>number</VALUE> <VALUE>org_base</VALUE> <VALUE>org_ending</VALUE> <VALUE>org_key</VALUE> <VALUE>org_pre</VALUE> <VALUE>organization</VALUE> <VALUE>organization_noun</VALUE> <VALUE>person_ending</VALUE> <VALUE>person_first</VALUE> <VALUE>person_full</VALUE> <VALUE>phone_prefix</VALUE> <VALUE>sport</VALUE> <VALUE>spur</VALUE> <VALUE>spur_ident</VALUE> <VALUE>stop</VALUE> <VALUE>surname</VALUE> <VALUE>time</VALUE> <VALUE>time_modifier</VALUE> <VALUE>time_unit</VALUE> <VALUE>title</VALUE> <VALUE>year</VALUE> </VALUES> <!-- Optional: if present marks the attribute used as CLASS Only one attribute can be marked as class --> </ATTRIBUTE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>POS_category(-1)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Token</TYPE> <!-- Optional: the feature name for the feature used to extract values for the attribute --> <FEATURE>category</FEATURE> <!-- The position relative to the instance annotation --> <POSITION>-1</POSITION> <!-- The list of permitted values. if present, marks a nominal attribute; if absent, the attribute is numeric (double) --> <VALUES> <!-- One permitted value --> <VALUE>NN</VALUE> <VALUE>NNP</VALUE> <VALUE>NNPS</VALUE> <VALUE>NNS</VALUE> <VALUE>NP</VALUE> <VALUE>NPS</VALUE> <VALUE>JJ</VALUE> <VALUE>JJR</VALUE> <VALUE>JJS</VALUE> <VALUE>JJSS</VALUE> <VALUE>RB</VALUE> <VALUE>RBR</VALUE> <VALUE>RBS</VALUE> <VALUE>VB</VALUE> <VALUE>VBD</VALUE> <VALUE>VBG</VALUE> <VALUE>VBN</VALUE> <VALUE>VBP</VALUE> <VALUE>VBZ</VALUE> <VALUE>FW</VALUE> <VALUE>CD</VALUE> <VALUE>CC</VALUE> <VALUE>DT</VALUE> <VALUE>EX</VALUE> <VALUE>IN</VALUE> <VALUE>LS</VALUE> <VALUE>MD</VALUE> <VALUE>PDT</VALUE> <VALUE>POS</VALUE> <VALUE>PP</VALUE> <VALUE>PRP</VALUE> <VALUE>PRP$</VALUE> <VALUE>PRPR$</VALUE> <VALUE>RP</VALUE> <VALUE>TO</VALUE> <VALUE>UH</VALUE> <VALUE>WDT</VALUE> <VALUE>WP</VALUE> <VALUE>WP$</VALUE> <VALUE>WRB</VALUE> <VALUE>SYM</VALUE> <VALUE>\"</VALUE> <VALUE>#</VALUE> <VALUE>$</VALUE> <VALUE>’</VALUE> <VALUE>(</VALUE> <VALUE>)</VALUE> <VALUE>,</VALUE> <VALUE>--</VALUE> <VALUE>-LRB-</VALUE> <VALUE>.</VALUE> <VALUE>’’</VALUE> <VALUE>:</VALUE> <VALUE>::</VALUE> <VALUE>‘</VALUE> </VALUES> <!-- Optional: if present marks the attribute used as CLASS Only one attribute can be marked as class --> </ATTRIBUTE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>POS_category(0)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Token</TYPE> <!-- Optional: the feature name for the feature used to extract values for the attribute --> <FEATURE>category</FEATURE> <!-- The position relative to the instance annotation --> <POSITION>0</POSITION> <!-- The list of permitted values. if present, marks a nominal attribute; if absent, the attribute is numeric (double) --> <VALUES> <!-- One permitted value --> <VALUE>NN</VALUE> <VALUE>NNP</VALUE> <VALUE>NNPS</VALUE> <VALUE>NNS</VALUE> <VALUE>NP</VALUE> <VALUE>NPS</VALUE> <VALUE>JJ</VALUE> <VALUE>JJR</VALUE> <VALUE>JJS</VALUE> <VALUE>JJSS</VALUE> <VALUE>RB</VALUE> <VALUE>RBR</VALUE> <VALUE>RBS</VALUE> <VALUE>VB</VALUE> <VALUE>VBD</VALUE> <VALUE>VBG</VALUE> <VALUE>VBN</VALUE> <VALUE>VBP</VALUE> <VALUE>VBZ</VALUE> <VALUE>FW</VALUE> <VALUE>CD</VALUE> <VALUE>CC</VALUE> <VALUE>DT</VALUE> <VALUE>EX</VALUE> <VALUE>IN</VALUE> <VALUE>LS</VALUE> <VALUE>MD</VALUE> <VALUE>PDT</VALUE> <VALUE>POS</VALUE> <VALUE>PP</VALUE> <VALUE>PRP</VALUE> <VALUE>PRP$</VALUE> <VALUE>PRPR$</VALUE> <VALUE>RP</VALUE> <VALUE>TO</VALUE> <VALUE>UH</VALUE> <VALUE>WDT</VALUE> <VALUE>WP</VALUE> <VALUE>WP$</VALUE> <VALUE>WRB</VALUE> <VALUE>SYM</VALUE> <VALUE>\"</VALUE> <VALUE>#</VALUE> <VALUE>$</VALUE> <VALUE>’</VALUE> <VALUE>(</VALUE> <VALUE>)</VALUE> <VALUE>,</VALUE> <VALUE>--</VALUE> <VALUE>-LRB-</VALUE> <VALUE>.</VALUE> <VALUE>’’</VALUE> <VALUE>:</VALUE> <VALUE>::</VALUE> <VALUE>‘</VALUE> </VALUES> <!-- Optional: if present marks the attribute used as CLASS Only one attribute can be marked as class --> </ATTRIBUTE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>POS_category(1)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Token</TYPE> <!-- Optional: the feature name for the feature used to extract values for the attribute --> <FEATURE>category</FEATURE> <!-- The position relative to the instance annotation --> <POSITION>1</POSITION> <!-- The list of permitted values. if present, marks a nominal attribute; if absent, the attribute is numeric (double) --> <VALUES> <!-- One permitted value --> <VALUE>NN</VALUE> <VALUE>NNP</VALUE> <VALUE>NNPS</VALUE> <VALUE>NNS</VALUE> <VALUE>NP</VALUE> <VALUE>NPS</VALUE> <VALUE>JJ</VALUE> <VALUE>JJR</VALUE> <VALUE>JJS</VALUE> <VALUE>JJSS</VALUE> <VALUE>RB</VALUE> <VALUE>RBR</VALUE> <VALUE>RBS</VALUE> <VALUE>VB</VALUE> <VALUE>VBD</VALUE> <VALUE>VBG</VALUE> <VALUE>VBN</VALUE> <VALUE>VBP</VALUE> <VALUE>VBZ</VALUE> <VALUE>FW</VALUE> <VALUE>CD</VALUE> <VALUE>CC</VALUE> <VALUE>DT</VALUE> <VALUE>EX</VALUE> <VALUE>IN</VALUE> <VALUE>LS</VALUE> <VALUE>MD</VALUE> <VALUE>PDT</VALUE> <VALUE>POS</VALUE> <VALUE>PP</VALUE> <VALUE>PRP</VALUE> <VALUE>PRP$</VALUE> <VALUE>PRPR$</VALUE> <VALUE>RP</VALUE> <VALUE>TO</VALUE> <VALUE>UH</VALUE> <VALUE>WDT</VALUE> <VALUE>WP</VALUE> <VALUE>WP$</VALUE> <VALUE>WRB</VALUE> <VALUE>SYM</VALUE> <VALUE>\"</VALUE> <VALUE>#</VALUE> <VALUE>$</VALUE> <VALUE>’</VALUE> <VALUE>(</VALUE> <VALUE>)</VALUE> <VALUE>,</VALUE> <VALUE>--</VALUE> <VALUE>-LRB-</VALUE> <VALUE>.</VALUE> <VALUE>’’</VALUE> <VALUE>:</VALUE> <VALUE>::</VALUE> <VALUE>‘</VALUE> </VALUES> <!-- Optional: if present marks the attribute used as CLASS Only one attribute can be marked as class --> </ATTRIBUTE> <ATTRIBUTE> <!-- The name given to the attribute --> <NAME>Entity(0)</NAME> <!-- The type of annotation used as attribute --> <TYPE>Entity</TYPE> <!-- The position relative to the instance annotation --> <POSITION>0</POSITION> <CLASS/> <!-- Optional: if present marks the attribute used as CLASS Only one attribute can be marked as class --> </ATTRIBUTE> </DATASET> <ENGINE> <WRAPPER>gate.creole.ml.weka.Wrapper</WRAPPER> <OPTIONS> <CLASSIFIER OPTIONS="-S -C 0.25 -B -M 2">weka.classifiers.trees.J48</CLASSIFIER> <CONFIDENCE-THRESHOLD>0.85</CONFIDENCE-THRESHOLD> </OPTIONS> </ENGINE> </ML-CONFIG> |
