#8601This paper considers the problem of automatic assessment oflocal coherence.
other,4-2-P05-1018,ak
coherence
</term>
. We present a novel
<term>
entity-based representation of discourse
</term>
which is inspired by
<term>
Centering
#8608We present a novelentity-based representation of discourse which is inspired by Centering Theory and can be computed automatically from raw text.
other,12-2-P05-1018,ak
discourse
</term>
which is inspired by
<term>
Centering Theory
</term>
and can be computed automatically
#8616We present a novel entity-based representation of discourse which is inspired byCentering Theory and can be computed automatically from raw text.
other,20-2-P05-1018,ak
can be computed automatically from
<term>
raw text
</term>
. We view
<term>
coherence assessment
#8624We present a novel entity-based representation of discourse which is inspired by Centering Theory and can be computed automatically fromraw text.
tech,2-3-P05-1018,ak
from
<term>
raw text
</term>
. We view
<term>
coherence assessment
</term>
as a
<term>
ranking learning problem
#8629We viewcoherence assessment as a ranking learning problem and show that the proposed discourse representation supports the effective learning of a ranking function.
tech,6-3-P05-1018,ak
<term>
coherence assessment
</term>
as a
<term>
ranking learning problem
</term>
and show that the proposed
<term>
discourse
#8633We view coherence assessment as aranking learning problem and show that the proposed discourse representation supports the effective learning of a ranking function.
other,14-3-P05-1018,ak
problem
</term>
and show that the proposed
<term>
discourse representation
</term>
supports the effective learning of
#8641We view coherence assessment as a ranking learning problem and show that the proposeddiscourse representation supports the effective learning of a ranking function.
tech,22-3-P05-1018,ak
supports the effective learning of a
<term>
ranking function
</term>
. Our experiments demonstrate that
#8649We view coherence assessment as a ranking learning problem and show that the proposed discourse representation supports the effective learning of aranking function.
measure(ment),10-4-P05-1018,ak
model achieves significantly higher
<term>
accuracy
</term>
than a state-of-the-art
<term>
coherence
#8662Our experiments demonstrate that the induced model achieves significantly higheraccuracy than a state-of-the-art coherence model.
other,14-4-P05-1018,ak
accuracy
</term>
than a state-of-the-art
<term>
coherence model
</term>
. This paper presents a
<term>
corpus
#8666Our experiments demonstrate that the induced model achieves significantly higher accuracy than a state-of-the-artcoherence model.