|Title:||Cross-Lingual Annotation Projection Models for Role-Semantic Information||Add Dissertation|
|Author:||Sebastian Pado||Update Dissertation|
|Email:||click here to access email|
|Institution:||Saarland University, Department of Computational Linguistics and Phonetics|
|Linguistic Subfield(s):||Computational Linguistics; Text/Corpus Linguistics;|
|Abstract:||Due to the high cost of manual annotation, resources with role-semantic
annotation exist only for a small number of languages, notably English.
This thesis addresses the resulting resource scarcity problem by
developing methods to induce role-semantic annotation for new languages
We address the induction task with annotation projection, a general
procedure to exchange linguistic information between aligned sentences in a
parallel corpus. Annotation projection is a knowledge-lean approach, and
thus applicable even to resource-poor languages. We evaluate our approach
by using FrameNet, a large English resource for frame semantics, to induce
frame-semantic annotation for two target languages, German and French.
We project semantic classes and roles in two separate steps, since the two
tasks have different profiles. The projection of semantic classes can be
realised using simply by using correspondences between predicates, which
are usually single words. Translational shifts, i.e., translations which
change the semantic class (frame) of the original predicate, can be
filtered out with knowledge-lean filtering mechanisms that rely on
In contrast, the projection of semantic roles relies mainly on clean
correspondences between sentential constituents (i.e.,role-bearing
phrases). We show that such correspondences can be obtained by formalising
the task as a graph matching problem that integrates knowledge about
syntactic bracketings. The resulting correspondences show a high precision
even for noisy input data from automatic shallow semantic parsing.
In sum, the results of this thesis indicate that the semantic
generalisations made by frame semantics carry over to a considerable degree
from English to other languages not only on the type, but also on the token
level. The projection methods we have developed can be applied to robustly
and automatically create frame-semantic resources for new languages.