Our objectives are met by giving clear definitions that determine the
<term>
projection
</term>
of
<term>
structures
</term>
from the
<term>
lexicon
</term>
, and identify
<term>
maximal projections
</term>
,
<term>
auxiliary trees
</term>
and
<term>
foot nodes
</term>
.
<term>
Corpus-based sense disambiguation methods
</term>
, like most other
<term>
statistical NLP approaches
</term>
, suffer from the problem of
<term>
data sparseness
</term>
.
#27712Our objectives are met by giving clear definitions that determine the projection of structures from the lexicon, and identify maximal projections, auxiliary trees and foot nodes.Corpus-based sense disambiguation methods, like most other statistical NLP approaches, suffer from the problem of data sparseness.
tech,8-1-P95-1025,ak
<term>
Corpus-based sense disambiguation methods
</term>
, like most other
<term>
statistical NLP approaches
</term>
, suffer from the problem of
<term>
data sparseness
</term>
.
#27720Corpus-based sense disambiguation methods, like most otherstatistical NLP approaches, suffer from the problem of data sparseness.
other,17-1-P95-1025,ak
<term>
Corpus-based sense disambiguation methods
</term>
, like most other
<term>
statistical NLP approaches
</term>
, suffer from the problem of
<term>
data sparseness
</term>
.
#27729Corpus-based sense disambiguation methods, like most other statistical NLP approaches, suffer from the problem ofdata sparseness.
lr,13-2-P95-1025,ak
In this paper , we describe an approach which overcomes this problem using
<term>
dictionary definitions
</term>
.
#27745In this paper, we describe an approach which overcomes this problem usingdictionary definitions.
other,2-3-P95-1025,ak
Using the
<term>
definition-based conceptual co-occurrence data
</term>
collected from the relatively small
<term>
Brown corpus
</term>
, our
<term>
sense disambiguation system
</term>
achieves an
<term>
average accuracy
</term>
comparable to
<term>
human performance
</term>
given the same
<term>
contextual information
</term>
.
#27750Using thedefinition-based conceptual co-occurrence data collected from the relatively small Brown corpus, our sense disambiguation system achieves an average accuracy comparable to human performance given the same contextual information.
lr-prod,11-3-P95-1025,ak
Using the
<term>
definition-based conceptual co-occurrence data
</term>
collected from the relatively small
<term>
Brown corpus
</term>
, our
<term>
sense disambiguation system
</term>
achieves an
<term>
average accuracy
</term>
comparable to
<term>
human performance
</term>
given the same
<term>
contextual information
</term>
.
#27759Using the definition-based conceptual co-occurrence data collected from the relatively smallBrown corpus, our sense disambiguation system achieves an average accuracy comparable to human performance given the same contextual information.
tech,15-3-P95-1025,ak
Using the
<term>
definition-based conceptual co-occurrence data
</term>
collected from the relatively small
<term>
Brown corpus
</term>
, our
<term>
sense disambiguation system
</term>
achieves an
<term>
average accuracy
</term>
comparable to
<term>
human performance
</term>
given the same
<term>
contextual information
</term>
.
#27763Using the definition-based conceptual co-occurrence data collected from the relatively small Brown corpus, oursense disambiguation system achieves an average accuracy comparable to human performance given the same contextual information.
measure(ment),20-3-P95-1025,ak
Using the
<term>
definition-based conceptual co-occurrence data
</term>
collected from the relatively small
<term>
Brown corpus
</term>
, our
<term>
sense disambiguation system
</term>
achieves an
<term>
average accuracy
</term>
comparable to
<term>
human performance
</term>
given the same
<term>
contextual information
</term>
.
#27768Using the definition-based conceptual co-occurrence data collected from the relatively small Brown corpus, our sense disambiguation system achieves anaverage accuracy comparable to human performance given the same contextual information.
measure(ment),24-3-P95-1025,ak
Using the
<term>
definition-based conceptual co-occurrence data
</term>
collected from the relatively small
<term>
Brown corpus
</term>
, our
<term>
sense disambiguation system
</term>
achieves an
<term>
average accuracy
</term>
comparable to
<term>
human performance
</term>
given the same
<term>
contextual information
</term>
.
#27772Using the definition-based conceptual co-occurrence data collected from the relatively small Brown corpus, our sense disambiguation system achieves an average accuracy comparable tohuman performance given the same contextual information.
other,29-3-P95-1025,ak
Using the
<term>
definition-based conceptual co-occurrence data
</term>
collected from the relatively small
<term>
Brown corpus
</term>
, our
<term>
sense disambiguation system
</term>
achieves an
<term>
average accuracy
</term>
comparable to
<term>
human performance
</term>
given the same
<term>
contextual information
</term>
.
#27777Using the definition-based conceptual co-occurrence data collected from the relatively small Brown corpus, our sense disambiguation system achieves an average accuracy comparable to human performance given the samecontextual information.