tailieunhanh - Báo cáo khoa học: "Edit Tree Distance alignments for Semantic Role Labelling"

Tree SRL system‖ is a Semantic Role Labelling supervised system based on a tree-distance algorithm and a simple k-NN implementation. The novelty of the system lies in comparing the sentences as tree structures with multiple relations instead of extracting vectors of features for each relation and classifying them. The system was tested with the English CoNLL-2009 shared task data set where 79% accuracy was obtained. | Edit Tree Distance alignments for Semantic Role Labelling Hector-Hugo Franco-Penya Trinity College Dublin Dublin Ireland. francoph@cs . Abstract Tree SRL system is a Semantic Role Labelling supervised system based on a tree-distance algorithm and a simple k-NN implementation. The novelty of the system lies in comparing the sentences as tree structures with multiple relations instead of extracting vectors of features for each relation and classifying them. The system was tested with the English CoNLL-2009 shared task data set where 79 accuracy was obtained. 1 Introduction Semantic Role Labelling SRL is a natural language processing task which deals with semantic analysis at sentence-level. SRL is the task of identifying arguments for a certain predicate and labelling them. The predicates are usually verbs. They establish what happened . The arguments determine events such as who whom where etc with reference to one predicate. The possible semantic roles are pre-defined for each predicate. The set of roles depends on the corpora. SRL is becoming an important tool for information extraction text summarization machine translation and question answering Màrquez et al 2008 . 2 The data The data set I used is taken from the CoNLL-2009 shared task Hajic et al. 2009 and is part of Propbank. Propbank Palmer et al 2005 is a hand-annotated corpus. It transforms sentences into propositions. It adds a semantic layer to the Penn TreeBank Marcus et al 1994 and defines a set of semantic roles for each predicate. It is difficult to define universal semantic roles for all predicates. That is why PropBank defines a set of semantic roles for each possible sense of each predicate frame See a sample of the frame raise on the Figure 1 caption . Sentences predicates arguments Predicates per sentence arguments per sub-tree File size in Mb Tra 39279 179014 393699 Dev 1334 6390 13865 Evl 2399 10498 23286 Table 1 The data The data set is .