<?xml version="1.0" standalone="yes"?>
<Paper uid="P05-3002">
  <Title>Accessing GermaNet Data and Computing Semantic Relatedness</Title>
  <Section position="3" start_page="5" end_page="5" type="intro">
    <SectionTitle>
2 GermaNet API
</SectionTitle>
    <Paragraph position="0"> The API for accessing GermaNet has to provide functions similar to the API developed for WordNet.</Paragraph>
    <Paragraph position="1"> We evaluated the C-library distributed together with GermaNet V4.0 and the XML encoded version of GermaNet (Lemnitzer &amp; Kunze, 2002). As we wanted the code to be portable across platforms, we built upon the latter. The XML version of GermaNet is parsed with the help of the Apache Xerces parser, http://xml.apache.org/ to create a JAVA object representing GermaNet. For stemming the words, we use the functionality provided by the Porter stemmer for the German language, freely available from http://snowball.tartarus.org/german/stemmer.html.</Paragraph>
    <Paragraph position="2"> Thus, the GermaNet object exists in two versions, the original one, where the information can be accessed using words, and the stemmed one, where the information can be accessed using word stems.</Paragraph>
    <Paragraph position="3"> We implemented a range of JAVA based methods for querying the data. These methods are organized around the notions of word sense and synset. On the word sense (WS) level, we have the following methods: getAntonyms() retrieves all antonyms of a given WS; getArtificial() indicates whether a WS is an artificial concept; getGrapheme() gets a graphemic representation of a WS; getParticipleOf() retrieves the WS of the verb that the word sense is a participle of; getPartOfSpeech() gets the part of speech associated with a WS; getPertonym() gives the WS that the word sense is derived from; get-ProperName() indicates whether the WS is a proper name; getSense() yields the sense number of a WS in GermaNet; getStyle() indicates if the WS is stylistically marked; getSynset() returns the corresponding synset; toString() yields a string representing a WS.</Paragraph>
    <Paragraph position="4"> On the synset level, the following information can be accessed: getAssociations() returns all associations; getCausations() gets the effects that a given synset is a cause of; getEntailments() yields synsets that entail a given synset; getHolonyms(), getHyponyms(), getHypernyms(), getMeronyms() return a list of holonyms, hyponyms, immediate hypernyms, and meronyms respectively; getPartOfSpeech() returns the part of speech associated with word senses of a synset; getWordSenses() returns all word senses constituting the synset; toString() yields a string representation of a synset.</Paragraph>
    <Paragraph position="5"> The metrics of semantic relatedness are designed to employ this API. They are implemented as classes which use the API methods on an instance of the GermaNet object.</Paragraph>
  </Section>
class="xml-element"></Paper>