Diagnosing Vocal Disorders using Cobweb Clustering of the Jitter, Shimmer, and Harmonics-to-Noise Ratio
- Authors
- Lee, Keonsoo; Moon, Chanki; Nam, Yunyoung
- Issue Date
- 30-Nov-2018
- Publisher
- 한국인터넷정보학회
- Keywords
- Cobweb; Clustering; HNR; Jitter; Shimmer; Vocal Disorder
- Citation
- KSII Transactions on Internet and Information Systems, v.12, no.11, pp 5541 - 5554
- Pages
- 14
- Journal Title
- KSII Transactions on Internet and Information Systems
- Volume
- 12
- Number
- 11
- Start Page
- 5541
- End Page
- 5554
- URI
- https://scholarworks.bwise.kr/sch/handle/2021.sw.sch/5490
- DOI
- 10.3837/tiis.2018.11.020
- ISSN
- 1976-7277
1976-7277
- Abstract
- A voice is one of the most significant non-verbal elements for communication. Disorders in vocal organs, or habitual muscular setting for articulatory cause vocal disorders. Therefore, by analyzing the vocal disorders, it is possible to predicate vocal diseases. In this paper, a method of predicting vocal disorders using the jitter, shimmer, and harmonics-to-noise ratio (HNR) extracted from vocal records is proposed. In order to extract jitter, shimmer, and HNR, one-second's voice signals are recorded in 44.1 khz. In an experiment, 151 voice records are collected. The collected data set is clustered using cobweb clustering method. 21 classes with 12 leaves are resulted from the data set. According to the semantics of jitter, shimmer, and HNR, the class whose centroid has lowest jitter and shimmer, and highest HNR becomes the normal vocal group. The risk of vocal disorders can be predicted by measuring the distance and direction between the centroids.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - College of Engineering > Department of Computer Science and Engineering > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/sch/handle/2021.sw.sch/5490)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.