Table 1 - uploaded by Antonio Ferrández
Content may be subject to copyright.
Applied English-Spanish contrastive grammar rules
Source publication
In QA@CLEF 2008, we participate in monolingual (Spanish) and multilingual (English - Spanish) tasks. Specifically, in this paper, we will tackle with the English - Spanish QA task. In this edition we will deal with two main problems: an heterogeneous document collection (news articles and Wikipedia) and a large number of topic-related questions, wh...
Context in source publication
Context 1
... applied English-Spanish contrastive grammar rules are based on the ones derived from the previous study developed by Fernandez et al. [2] and Martinez-Vazquez [6]. The applied English- Spanish contrastive grammar rules are detailed in Table 1, and an example of translation, applying each one of these rules, is explained in Finally, the translation performed by the rest of predicates of the logic form, whose logic struc- ture does not match these English-Spanish contrastive grammar rules, consists in the concatenation of the sequence of translations of these predicates. ...
Similar publications
We describe our system for the monolingual Dutch and multilingual English to Dutch QA tasks. First, we present a brief overview of our QA-system, which makes heavy use of syntactic information. Next, we describe the modules that were developed especially for CLEF 2007, i.e. preprocessing of Wikipedia, inclusion of query expansion in IR, anaphora re...
This paper presents our system for the CoNLL 2009 Shared Task on Syntactic and Semantic Dependencies in Multiple Languages (Hajic et al., 2009). In this work we focus only on the Semantic Role Labelling (SRL) task. We use Markov Logic to define a joint SRL model and achieve the third best average performance in the closed Track for SRLOnly systems...
The theory of Human Plausible Reasoning (HPR) is an attempt by Collins and Michalski to explain how people answer questions
when they are uncertain. The theory consists of a set of patterns and a set of inferences which could be applied on those
patterns. This paper, investigates the application of HPR theory to the domain of cross language filteri...
This paper presents a robust and efficient text detection algorithm for news video. The proposed algorithm uses the temporal information of video and logical AND operation to remove most of irrelevant background. Then a window-based method by counting the black-and-white transitions is applied on the resulted edge map to obtain rough text blobs. Li...
Establishing correspondences between wordnets of different languages is essential to multilingual knowledge processing. We claim that such correspon- dences must be based on lexical semantic relations, rather than top ontology or word translations. In particular, we define a translation equivalence relation as a bilingual lexical semantic relation....
Citations
... It is essential in MT evaluation and, in the current cross-language setting, to identify parallel corpora to feed a (neural) machine translation model with (Munteanu and Marcu, 2005). Efforts have been carried out to approach crosslanguage versions of these tasks without translating all the texts into one common language (e.g., (Bouma et al., 2008;Muñoz Terol et al., 2008;Potthast et al., 2011)), but using interlingua or multilingual representations instead. Still, such representations are usually hard to design and this is precisely when our neural context vector NMT embedding representation comes into play. ...
End-to-end neural machine translation has overtaken statistical machine translation in terms of translation quality for some language pairs, specially those with a large amount of parallel data available. Beside this palpable improvement, neural networks embrace several new properties. A single system can be trained to translate between many languages at almost no additional cost other than training time. Furthermore, internal representations learned by the network serve as a new semantic representation of words -or sentences- which, unlike standard word embeddings, are learned in an essentially bilingual or even multilingual context. In view of these properties, the contribution of the present work is two-fold. First, we systematically study the context vectors, i.e. output of the encoder, and their prowess as an interlingua representation of a sentence. Their quality and effectiveness are assessed by similarity measures across translations, semantically related, and semantically unrelated sentence pairs. Second, and as extrinsic evaluation of the first point, we identify parallel sentences in comparable corpora, obtaining an F1=98.2% on data from a shared task when using only context vectors. F1 reaches 98.9% when complementary similarity measures are used.
