Speaker Classification by Means of Orthographic and Broad Phonetic Transcriptions of Speech

Bael, Christophe Van; Halteren, H. van

doi:10.1007/978-3-540-74122-0_22

Cited by 1 publication

(6 citation statements)

References 9 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Van Bael and Van Halteren (2007) investigated classification of groups of speakers on the basis of phonetic transcriptions and reported that classification based on pronunciation features performed poorly. The authors discussed several possible explanations for this finding.…”

Section: Discussionmentioning

confidence: 99%

“…Similar to Van Bael and Van Halteren (2007), to study differences in the choice and reduction of words between individual speakers, we applied a classification algorithm, in which speech fragments are attributed to their speakers on the basis of lexical and pronunciation patterns. If classification results in high performance scores, this would indicate that speakers differ in their speech habits.…”

Section: Introductionmentioning

confidence: 99%

“…Van Bael and Van Halteren (2007) studied the effects of the speaker's age, gender, regional background, and level of education on word choice and pronunciation variation by classifying speakers belonging to groups differing in these characteristics. Using automatically generated broad phonetic transcriptions of the telephone dialogues of the Spoken Dutch Corpus (Oostdijk, 2002), the authors generated two sets of classification features: one set of approximately 150,000 lexical features, including average utterance length, part-of-speech tags, and uni-, bi-, and trigram counts, and another set of 94 pronunciation features representing phone differences between full pronunciations of words and their actual phonetic transcriptions.…”

Section: Introductionmentioning

confidence: 99%

“…The authors suggested that this may be due to the broadness of the phonetic transcriptions, the limited set of pronunciation features, or the heterogeneity within their speaker groups. This paper builds on the work by Van Bael and Van Halteren (2007) by also using a classifier to investigate pronunciation differences between speakers and by applying the classifier on phonetic transcriptions of conversations instead of the speech signal itself. However, we address related but different research questions, since we focus on Dutch speakers who have the same regional background, gender, and educational level; that is, a homogeneous set of speakers.…”

Section: Introductionmentioning

confidence: 99%

“…We investigate whether these speakers show individual differences in their choice and pronunciation of words in casual conversations, and if so, how these speakers differ. In addition, our study differs from Van Bael and Van Halteren (2007) in that we use a different classifier and different features, and we developed our own research approach.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Choice and pronunciation of words: Individual differences within a homogeneous group of speakers

Hanique¹,

Ernestus

Boves³

2015

Corpus Linguistics and Linguistic Theory

View full text Add to dashboard Cite

This paper investigates whether individual speakers forming a homogeneous group differ in their choice and pronunciation of words when engaged in casual conversation, and if so, how they differ. More specifically, it examines whether the Balanced Winnow classifier is able to distinguish between the twenty speakers of the Ernestus Corpus of Spontaneous Dutch, who all have the same social background. To examine differences in choice and pronunciation of words, instead of characteristics of the speech signal itself, classification was based on lexical and pronunciation features extracted from hand-made orthographic and automatically generated broad phonetic transcriptions. The lexical features consisted of words and two-word combinations. The pronunciation features represented pronunciation variations at the word and phone level that are typical for casual speech. The best classifier achieved a performance of 79.9% and was based on the lexical features and on the pronunciation features representing single phones and triphones. The speakers must thus differ from each other in these features. Inspection of the relevant features indicated that, among other things, the words relevant for classification generally do not contain much semantic content, and that speakers differ not only from each other in the use of these words but also in their pronunciation.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%