This paper outlines the first phase of the PRAGMATEXT project. The aim of PRAGMATEXT is to introduce pragmatic knowledge into the transcriptions of C- ORAL-ROM, a spontaneous spoken corpus of Spanish. The paper is divided in four sections. The first section presents the most relevant features of the C-ORAL-ROM corpus. The second describes the pragmatic-discursive
Read more Textual typology and transcription problems: analysis of the data In this section, the results obtained by the program are analyzed. The analysis procedure has always been the same for each of the linguistic phenomena studied: First, the relation between frequency of occurrence of the features and textual typology is stated. Second, we evaluated whether this
Read more The computational tool The C-ORAL-ROM corpus is tagged with XML. Using the information included in the tags, we developed a program which automatically calculate the frequency of occurrence of each of the following features: overlapping, retracting, number of dialogic turns, speaking speed, fragmented words and supports. These frequencies were calculated for each class of texts.
Read more Orality and transcription problems: the original hypothesis In order to find out what kind of relation there is between orality and linguistic registers, two scales of transcription difficulty were stated taking into consideration the following two parameters: Degree of formality (Scale 1) Two ends can be considered when dealing with the texts in terms of
Read more The notion of orality It is well known that spoken language is not always a synonym to orality, if we understand orality as the presence of linguistic, paralinguistic and interactive phenomena, such as retracting or overlapping, which are not present in the written register. The registers in spoken language vary depending on the communicative situation.
Read more Description of the corpus C-ORAL-ROM is a multilingual spontaneous speech corpus (Cresti et al., 2002) of the four main roman languages: French, Italian, Portuguese and Spanish. Each subcorpus consists of around 300,000 words. With the aim of enabling comparability between the different subcorpora, several sampling criteria concerning the distribution of the corpus were established: as
Read more 1. Introduction This paper is the result of an experiment carried out by a group of transcribers at the Laboratorio de Lingüística Informática (LLI) at the Universidad Autónoma de Madrid, once the recording and transcribing phases of the C-ORAL-ROM project were over. The goal of the experiment was to confirm certain hypothesis which had arisen
Read more Abstract This paper analyses the effects of certain oral features on the process of transcription of spontaneous speech recordings. On the basis of the statistical analysis of the data obtained from the C-ORAL-ROM corpus, it will be shown empirically that transcription difficulties vary according to the communicative situation, the degree of formality and the number
Read more Conclusions and Future Work The results are approximately 60 % failures and 40 % success. Behind these figures lay the other’s image and the expectations, both from the speaker and from the machine. One of the principles that makes human communication possible is that speakers are able to adapt their speech acts to the expectations
Read more Analysing the interactions System evaluation The first step had been to measure the degree of real success of the dialog system. As the phone calls were intended only for scientific purposes and the speakers proposed fictitious destinations, it is not easy to define what “success” mean. As we will see later, some speakers change their
Read more The man-machine telephone corpus For this paper, we have only considered the small subset of human-machine telephone recordings, part of the telephone corpus. This subset has been collected by means of an automatic telephone call system specifically designed and developed for the C-ORAL- ROM project by ITC-Irst (Falavigna & Gretter 2003). The system handles interactions
Read more Introducción The goal of this paper is to analyse man-machine interactions in the Spanish data of C- ORAL-ROM, a multilingual, multimedia corpus of spontaneous speech. The subcorpus consists of 41 recordings, over 2600 words. The issue we want to address is why some of the interactions succeeded, while most of them resulted in misunderstandings and
Read moreAna Gonzalez Ledesma ©. All rights reserved.
