On the one hand, the authors confirm that the use of only machine-learning methods is highly dependent on the annotated training data, and thus obtained better results for well-represented classes. On the other hand, the use of only a rule-based method was not sufficient to deal with new types of data. Finally, the use of hybrid approaches combining machine-learning and rule-based approaches yielded higher scores.
The various meanings of discourse connectives like while and however are difficult
to identify and annotate, even for trained human annotators. This problem is all the
more important that connectives are salient textual markers of cohesion and need to be
correctly interpreted for many NLP applications. In this paper, we suggest an
alternative route to reach a reliable annotation of connectives, by making use of the
information provided by their translation in large parallel corpora. This method thus
replaces the difficult explicit reasoning involved in traditional sense annotation by an
empirical clustering of the senses emerging from the translations. We argue that this
method has the advantage of providing more reliable reference data than traditional
sense annotation. In addition, its simplicity allows for the rapid constitution of large
annotated datasets.
Discourse connectives are often said to be language specific, and therefore not easily paired with a translation equivalent in a target language. However, few studies have assessed the magnitude and the causes of these divergences. In this paper, we provide an overview of the similarities and discrepancies between causal connectives in two typologically related languages: English and French. We first discuss two criteria used in the literature to account for these differences: the notion of domains of use and the information status of the cause segment. We then test the validity of these criteria through an empirical contrastive study of causal connectives in English and French, performed on a bidirectional corpus. Our results indicate that French and English connectives have only partially overlapping profiles and that translation equivalents are adequately predicted by these two criteria.
The search for translation universals has been an important topic in translation studies over the past decades. In this paper, we focus on the notion of explicitation through a multifaceted study of causal connectives, integrating four different variables: the role of the source and the target languages, the influence of specific connectives and the role of the discourse relation they convey. Our results indicate that while source and target languages do not globally influence explicitation, specific connectives have a significant impact on this phenomenon. We also show that in English and French, the most frequently used connectives for explicitation share a similar semantic profile. Finally, we demonstrate that explicitation also varies across different discourse relations, even when they are conveyed by a single connective.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.