scholarly journals Supertagging Combinatory Categorial Grammar with Attentive Graph Convolutional Networks

Author(s):  
Yuanhe Tian ◽  
Yan Song ◽  
Fei Xia
2007 ◽  
Vol 33 (3) ◽  
pp. 355-396 ◽  
Author(s):  
Julia Hockenmaier ◽  
Mark Steedman

This article presents an algorithm for translating the Penn Treebank into a corpus of Combinatory Categorial Grammar (CCG) derivations augmented with local and long-range word-word dependencies. The resulting corpus, CCGbank, includes 99.4% of the sentences in the Penn Treebank. It is available from the Linguistic Data Consortium, and has been used to train wide-coverage statistical parsers that obtain state-of-the-art rates of dependency recovery. In order to obtain linguistically adequate CCG analyses, and to eliminate noise and inconsistencies in the original annotation, an extensive analysis of the constructions and annotations in the Penn Treebank was called for, and a substantial number of changes to the Treebank were necessary. We discuss the implications of our findings for the extraction of other linguistically expressive grammars from the Treebank, and for the design of future treebanks.


2004 ◽  
Vol 63 ◽  
Author(s):  
Rodrigo Tadeu Gonçalves ◽  
Luiz Arthur Pagani

O presente artigo trata das chamadas sentenças-labirinto, mostrando que, em casos em que a entoação e a estrutura informacional são suficientemente claras, a ambigüidade gerada pelo mencionado efeito não ocorre. O artigo contribui para a área do processamento lingüístico humano mostrando que, quando faladas, as sentenças das quais se esperam problemas de processamento sérios podem não apresentar tais problemas. A partir de um modelo teórico chamado Gramática Categorial Combinatória, mostramos como o processamento incremental de sentenças é ajudado pelas informações prosódicas e informacionais na atribuição de estrutura gramatical adequada a sentenças tradicionalmente consideradas “labirinto”. Garden-path effect beyond syntax: eliminating ambiguity Abstract The present article deals with the so-called garden-path effect. Traditionally, garden-path sentences are those that cause serious problem for the mental parser during processing and, although they are perfectly grammatical, there is no attribution of grammatical structure to them. We try to show that, when spoken, the garden-path sentences may not present the same kind of problem to the human sentence processing mechanism. In this paper we show how sufficiently informative data regarding prosody and informational structure can help the parser attribute correct grammatical structure to garden-path sentences when they are spoken. Using a framework called Combinatory Categorial Grammar, we show how incremental interpretation of garden-path sentences can be helped by prosody and informational structure during the processing of such sentences.


Sign in / Sign up

Export Citation Format

Share Document