tech,4-1-P03-1058,ak of interest . A central problem of <term> word sense disambiguation ( WSD ) </term> is the lack of <term> manually sense-tagged
lr,14-1-P03-1058,ak disambiguation ( WSD ) </term> is the lack of <term> manually sense-tagged data </term> required for <term> supervised learning
tech,19-1-P03-1058,ak sense-tagged data </term> required for <term> supervised learning </term> . In this paper , we evaluate an
lr,11-2-P03-1058,ak approach to automatically acquire <term> sense-tagged training data </term> from <term> English-Chinese parallel
lr,15-2-P03-1058,ak sense-tagged training data </term> from <term> English-Chinese parallel corpora </term> , which are then used for disambiguating
other,26-2-P03-1058,ak are then used for disambiguating the <term> nouns </term> in the <term> SENSEVAL-2 English lexical
other,29-2-P03-1058,ak disambiguating the <term> nouns </term> in the <term> SENSEVAL-2 English lexical sample task </term> . Our investigation reveals that
lr,8-3-P03-1058,ak reveals that this method of acquiring <term> sense-tagged data </term> is promising . On a subset of the
other,7-4-P03-1058,ak On a subset of the most difficult <term> SENSEVAL-2 nouns </term> , the <term> accuracy difference </term>
measure(ment),11-4-P03-1058,ak <term> SENSEVAL-2 nouns </term> , the <term> accuracy difference </term> between the two approaches is only
lr,37-4-P03-1058,ak if we disregard the advantage that <term> manually sense-tagged data </term> have in their sense coverage . Our
other,10-5-P03-1058,ak highlights the importance of the issue of <term> domain dependence </term> in evaluating <term> WSD programs </term>
tech,14-5-P03-1058,ak domain dependence </term> in evaluating <term> WSD programs </term> . We describe the ongoing construction
skrij podrobnosti