• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.9
• /
• pp.4264-4279
• /
• 2017

In this paper, maximum likelihood-based automatic lexicon generation using mixed-syllables is proposed for unlimited vocabulary voice interface for East Asian languages (e.g. Korean, Chinese and Japanese) in AI-assistant based interaction with mobile devices. The conventional lexicon has two inevitable problems: 1) a tedious repetition of out-of-lexicon unit additions to the lexicon, and 2) the propagation of errors during a morpheme analysis and space segmentation. The proposed method provides an automatic framework to solve the above problems. The proposed method produces a level of overall accuracy similar to one of previous methods in the presence of one out-of-lexicon word in a sentence, but the proposed method provides superior results with the absolute improvements of 1.62%, 5.58%, and 10.09% in terms of word accuracy when the number of out-of-lexicon words in a sentence was two, three and four, respectively.

• Proceedings of the IEEK Conference
• /
• summer
• /
• pp.391-394
• /
• 2004

This paper focuses on the DSP implementation of an HMM-based speech recognizer that can handle several hundred words of vocabulary size as well as speaker independency. First, we develop an HMM-based speech recognition system on the PC that operates on the frame basis with parallel processing of feature extraction and Viterbi decoding to make the processing delay as small as possible. Many techniques such as linear discriminant analysis, state-based Gaussian selection, and phonetic tied mixture model are employed for reduction of computational burden and memory size. The system is then properly optimized and compiled on the TMS320C6711 DSP for real-time operation. The implemented system uses 486kbytes of memory for data and acoustic models, and 24.5kbytes for program code. Maximum required time of 29.2ms for processing a frame of 32ms of speech validates real-time operation of the implemented system.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.33-36
• /
• 2000

어휘독립 고립단어인식은 미리 훈련된 부단어(sub-word) 단위의 음향모델을 이용하여 수시로 변하는 인식대상어휘를 인식하는 것이다. 본 논문에서는 소용량 음성 데이터베이스를 이용하여 어휘독립음성인식 시스템을 구성하였다. 소용량 음성 데이터베이스에서 미관측문맥 종속형 부단어에 대한 처리에 효과적인 백오프 기법을 이용한 음소 군집화 방법으로 문턱값을 변화시키며 인식실험을 수행하였다. 그리고 훈련용 데이터의 부족으로 인하여 문맥 종속형 부단어 모델이 훈련용 데이터베이스로 편중되는 문제를 deleted interpolation 방법을 이용하여 문맥 종속형 부단어 모델과 문맥 독립형 부단어 모델을 병합함으로써 해결하였다. 그 결과 음성인식의 성능이 향상되었다.

• Korean Linguistics
• /
• v.76
• /
• pp.83-112
• /
• 2017

In this paper, newspaper articles were collected through '5.18' keyword search results and the news corpus was constructed from the collected data. In the articles of local newspapers 'Honam' and 'Yeongnam', the ideological differences regarding '5.18' were investigated. The ideological differences of local newspaper discourse through objective figures was analyzed.. The subjects of the newspaper articles, the frequency of nouns and predicates were analyzed. The use and meaning of the intended vocabulary were examined. As a result of analyzing the title of the newspaper article, the discourse written in 'Honam' emphasized the necessity of re - recognition of 5.18. In both regions, the word "Gwangju" is often used. However, 'Gwangju' in 'Honam' newspaper means spiritual space, not physical space. In Honam regional newspapers, there are many vocabularies describing the events such as 'shoot' and 'fire', this calls for recollection and memory of '5.18'. In the analysis of newspaper discourse, the analysis of the contrast between the local newspapers was very insignificant, but, this study was conducted to analyze the discourse among local newspapers.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.5
• /
• pp.21-27
• /
• 1999

In this paper, we propose an improved MLP-based English pronunciation dictionary generator to apply to the variable vocabulary word recognizer. The variable vocabulary word recognizer can process any words specified in Korean word lexicon dynamically determined according to the current recognition task. To extend the ability of the system to task for English words, it is necessary to build a pronunciation dictionary generator to be able to process words not included in a predefined lexicon, such as proper nouns. In order to build the English pronunciation dictionary generator, we use context-dependent grapheme-to-phoneme multi-layer perceptron(MLP) architecture for each grapheme. To train each MLP, it is necessary to obtain grapheme-to-phoneme training data from general pronunciation dictionary. To automate the process, we use dynamic programming(DP) algorithm with some distance metrics. For training and testing the grapheme-to-phoneme MLPs, we use general English pronunciation dictionary with about 110 thousand words. With 26 MLPs each having 30 to 50 hidden nodes and the exception grapheme lexicon, we obtained the word accuracy of 72.8% for the 110 thousand words superior to rule-based method showing the word accuracy of 24.0%.

• Science of Emotion and Sensibility
• /
• v.12 no.1
• /
• pp.97-108
• /
• 2009

In the process of product design, shaping is the process of making a substantive existence, and ultimately it generates the outcome. The process of shaping is generally led by designer's initiative work, and in this process, various formative elements are used to generate the outcome. In this research, the basic purposes are to figure out the differences of elements which generated by the differences of consumer's and designer's view in the process of shaping of the product, and the characteristics of the affective responses caused by those differences. Also, it will examine how the consumers can directly participate in the process of the shaping of the consumer-participated product, and the feasible guidelines of design in which consumers' needs can be reflected more efficiently to the process of shaping. As a result, consumers and designers have certain degree of difference of view-point about the formative element of the shape. The difference was due to subjective common ideas of design in case of designers, and in case of consumers, it was due to their immature visual understanding. There is another experiment of affective response about the shape of the product. First, I established the sensible image vocabulary based on the shape of the product. And based on the vocabulary, I carried out the same experiments to the consumers and designers.

• 재활복지
• /
• v.21 no.2
• /
• pp.233-246
• /
• 2017

The present study is a basic study for application of Augmented Reality (AR) to verb teaching for children with language developmental disorders and is intended to examine validity for the list of verbs at the beginning of development. To confirm the validity of the verbs list, the appropriateness of the verbs was evaluated by three professors with certification of KSLP (Korean Speech-Language Pathologist) working in the department of Speech-Language Pathology at the university. The motion validity test was conducted by showing motion implemented as AR to eight master's students in Speech-Language Pathology major, having them record verbs that came to their mind, and evaluating in the conformity. The second motion validity test was conducted by using 5-point Likert scales to 87 undergraduates in Speech-Language Pathology major and having them see the motions in AR and marked the degrees to which them see the motions conform to the relevant verbs on the scales. Using the SPSS 21.0 program, descriptive statics analyses of the results were conducted. Through this all process, thirty verbs were selected as having content validity. It could be seen that when AR based communication system are applied, things and backgrounds that complement the insufficient movements of motions and help motion recognition should be also provided. In future studies, the 3D images of the AR based communication system will be complemented and the content validity will be verified with typically developing children and the children with language developmental disorders.

• Journal of Digital Convergence
• /
• v.20 no.1
• /
• pp.213-220
• /
• 2022

The purpose of this study is to investigate the relationship between the characteristics of using new words, ability to use new words, and peer popularity among students with mild intellectual disabilities and general students of the age-matched group. A total of 8 students, 4 students with mild intellectual disabilities aged 14 to 16 years of age and 4 normal students in the age-matched group, were compared between groups using a nonparametric test. In the case of new words, 60 new words were selected through expert content validity among 301 new words, and then recognition and background information on the 60 new words were collected. As a result of the study, first, there was a significant difference in understanding of new words between the student group with mild intellectual disability and the general student group of the same age. Second, the correlation between the use of new words and the popularity of peers was compared for each group of students with mild intellectual disabilities and a group of general students of the same age as possible. Therefore, when providing vocabulary instruction for students with mild intellectual disabilities, it suggests that it is necessary to teach new vocabulary in order to increase their relationship with their peers and their popularity.

• Journal of KIISE:Software and Applications
• /
• v.35 no.6
• /
• pp.386-391
• /
• 2008

Some spoken word errors that violate grammatical or writing rules occurs frequently in communication environments like mobile phone and messenger. These unexpected errors cause a problem in a language processing system for many applications like speech recognition, text-to-speech translation, and so on. In this paper, we proposed and implemented an automatic correction system of ill-formed words and word spacing errors in SMS sentences that has been the major errors of poor accuracy. We experimented three methods of constructing the word correction dictionary and evaluated the results of those methods. They are (1) manual construction of error words from the vocabulary list of ill-formed communication languages, (2) automatic construction of error dictionary from the manually constructed corpus, and (3) context-dependent method of automatic construction of error dictionary.

• The Journal of the Korea Contents Association
• /
• v.12 no.10
• /
• pp.541-554
• /
• 2012

This study suggests the effectiveness of G-learning English afterschool classes implemented to elementary school students at low achievement level in English. These days, the use of games in teaching and learning, known as G-learning, has gradually expanded, so it is necessary to consider how to adapt G-learning generally in English education. A G-learning afterschool English class was implemented to 23 low-level 5th grade students in an elementary school located in Daejon for 12 weeks. This study set two hypotheses aiming to determine the effectiveness in achievement and affectiveness of the participants. Pre and post achievement tests were conducted. Also, survey and FGI (focused group interview) were carried out twice with the participants. The study found that students' spelling awareness, vocabulary recognition and dialogue comprehension ability (hypothesis 1) were all improved with statistical significance. Moreover, after the class, participants' confidence and interest toward English study showed meaningful increases.