Event Extraction from Unstructured Amharic Text

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 195–203 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 195 Using the RUPEX Multichannel Corpus in a Pilot fMRI Study on Speech Disfluencies Katerina V. Smirnova1, Nikolay A. Korotaev2, Yana R. Panikratova3, Irina S. Lebedeva3, Ekaterina V. Pechenkova4, Olga V. Fedorova1 1Lomonosov Moscow State University, 2Russian State University for the Humanities, 3Mental Health Research Center, 4National Research University Higher School of Economics 1Leninskie Gory 1, 2Miusskaya sq. 6, 3Kashirskoye sh. 34, 4Armyanskiy ln. 4, bld. 2, Moscow, Russia kategold1@gmail.com, n_korotaev@hotmail.com, panikratova@mail.ru, irina.lebedeva@ncpz.ru, evp@virtualcoglab.org, olga.fedorova@msu.ru Abstract In modern linguistics and psycholinguistics speech disfluencies in real fluent speech are a well-known phenomenon. But it’s not still clear which components of brain systems are involved into its comprehension in a listener’s brain. In this paper we provide a pilot neuroimaging study of the possible neural correlates of speech disfluencies perception, using a combination of the corpus and functional magnetic-resonance imaging (fMRI) methods. Special technical procedure of selecting stimulus material from Russian multichannel corpus RUPEX allowed to create fragments in terms of requirements for the fMRI BOLD temporal resolution. They contain isolated speech disfluencies and their clusters. Also, we used the referential task for participants fMRI scanning. As a result, it was demonstrated that annotated multichannel corpora like RUPEX can be an important resource for experimental research in interdisciplinary fields. Thus, different aspects of communication can be explored through the prism of brain activation. Keywords: multichannel corpus, speech disfluency, functional magnetic-resonance imaging. 1. Introduction In natural communication, people often encounter problems in their discourse production. This may result in speech disfluencies, that is, various deviations from the “ideal”, fluent delivery (Clark and Clark, 1977). Numerous classifications of disfluencies have been proposed and / or implemented in annotating speech corpora, see, Maclay and Osgood, 1959; Levelt, 1993; Schriberg, 1994; Eklund, 2004, inter alia. They are mostly based on the speaker’s perspective. When planning their production, speakers may hesitate on a better way to express their intentions and search for a verbalization they will not have to reject afterwards. Alternatively, speakers may find an already- uttered discourse fragment inappropriate, stop their speech production “and then abort, recast or redo” their utterances, thus realizing self-repair (Fox et al., 2009: 59). Combined types of disfluencies are also possible, as will be shown in Section 3. Still, speech disfluencies also influence perception, as listeners may interpret them as signals of planning difficulties, possible places for turn-taking, and so on (see, e.g., Fox Tree, 2001; Barr and Seyfeddinipur, 2010). In this paper, we address various types of speech disfluencies from yet another perspective. We conducted a pilot event-related functional magnetic-resonance imaging (fMRI) study to reveal the neural correlates of disfluency perception in the listener’s brain. This research comes in line with the interdisciplinary initiative for analyzing multichannel discourse (see Kibrik and Fedorova, 2018a, b). We use data from the “Russian Pears Chats and Stories” (RUPEX) corpus, which was created as a resource …

Content truncated to 500 words