We propose in this paper a supervised learning approach to identify discourse relations in Arabic texts. To our knowledge, this work represents the first attempt to focus on both explicit and implicit relations that link adjacent as well as non adjacent Elementary Discourse Units (EDUs) within the Segmented Discourse Representation Theory (SDRT). We use the Discourse Arabic Treebank corpus (D-ATB) which is composed of newspaper documents extracted from the syntactically annotated Arabic Treebank v3.2 part3 where each document is associated with complete discourse graph according to the cognitive principles of SDRT. Our list of discourse relations is composed of a three-level hierarchy of 24 relations grouped into 4 top-level classes. To automatically learn them, we use state of the art features whose efficiency has been empirically proved. We investigate how each feature contributes to the learning process. We report our experiments on identifying fine-grained discourse relations, mid-level classes and also top-level classes. We compare our approach with three baselines that are based on the most frequent relation, discourse connectives and the features used by Al-Saif and Markert (2011). Our results are very encouraging and outperform all the baselines with an F-score of 78.1% and an accuracy of 80.6%. ª 2014 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
International audienceIn this article, we propose the first work that investigates the feasibility of Arabic discourse segmentation into elementary discourse units within the segmented discourse representation theory framework. We first describe our annotation scheme that defines a set of principles to guide the segmentation process. Two corpora have been annotated according to this scheme: elementary school textbooks and newspaper documents extracted from the syntactically annotated Arabic Treebank. Then, we propose a multiclass supervised learning approach that predicts nested units. Our approach uses a combination of punctuation, morphological, lexical, and shallow syntactic features. We investigate how each feature contributes to the learning process. We show that an extensive morphological analysis is crucial to achieve good results in both corpora. In addition, we show that adding chunks does not boost the performance of our system
We present in this paper an automatic summarization technique of Arabic texts, based on RST. We first present a corpus study which enabled us to specify, following empirical observations, a set of relations and rhetorical frames. Then, we present our method to automatically summarize Arabic texts. Finally, we present the architecture of the ARSTResume system. Our method is based on the Rhetorical Structure Theory (Mann, 1988) and uses linguistic knowledge. It relies on three pillars. The first consists in locating the rhetorical relations between the minimal units of the text by applying rhetorical rules. One of these units is the nucleus (the segment necessary to maintain coherence) and the other can be either nucleus or satellite (an optional segment). The second pillar is the representation and the simplification of the RST-tree that represents the source text in hierarchical form. The third pillar is the selection of sentences for the final summary, which takes into account the type of the rhetorical relations chosen for the extract. 2 LINGUISTIC ANALYSIS OF THE STUDY CORPUS An automatic summary requires, as a preliminary step, a linguistic analysis of the corpus (newspaper articles). The main goal of this analysis is to determine the surface linguistic units which represent linguistic markers as well as their corresponding validation markers. These linguistic markers are independent from a particular field and are organized in rhetorical relations.
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.