• 말소리와 음성과학
• /
• 제7권4호
• /
• pp.3-8
• /
• 2015

In this paper, we investigate an input dimension reduction method using continuous word vector in deep neural network language model. In the proposed method, continuous word vectors were generated by using Google's Word2Vec from a large training corpus to satisfy distributional hypothesis. 1-of-${\left|V\right|}$ coding discrete word vectors were replaced with their corresponding continuous word vectors. In our implementation, the input dimension was successfully reduced from 20,000 to 600 when a tri-gram language model is used with a vocabulary of 20,000 words. The total amount of time in training was reduced from 30 days to 14 days for Wall Street Journal training corpus (corpus length: 37M words).

• 말소리와 음성과학
• /
• 제8권4호
• /
• pp.9-14
• /
• 2016

This study investigated the patterns of coda deletion in spontaneous Seoul Korean speech. More specifically, the current study focused on three factors in promoting coda deletion, namely, word position, consonant type, and morpheme type. The results revealed that, first, coda deletion frequently occurred when affixes were attached to the ends of words, rather than in affixes in word-internal positions or in roots. Second, alveolar consonants [n] and [l] in the coda positions of high-frequency affixes [nɨn] and [lɨl] were most likely to be deleted. Additionally, regarding affix reduction in the word-final position, all subjects seemed to depend on this articulatory strategy to a similar degree. In sum, the current study found that affixes without primary semantic content in spontaneous speech tend to undergo the process of reduction, favoring the occurrence of specific pronunciation variants.

• 음성과학
• /
• 제13권3호
• /
• pp.35-48
• /
• 2006

The present study examined whether words with higher frequency have more exposure to the lenition process such as intervocalic stop voicing or /h/ reduction in the production of the Korean speakers. Experiment 1 and Experiment 2 tested if word-internal intervocalic voicing and /h/ reduction occur more often in the words with higher frequency than less frequent words respectively. Results showed that the rate of voicing was not significantly different between the high frequency group and the low frequency group; rather both high and low frequency words were shown to be fully voiced in this prosodic position. However, intervocalic /h/s were deleted more in high frequency words than in low frequency words. Low frequency words showed that other phonetic variants such as [h] and [w] were found more often than in high frequency group. Thus the results of the present study are indefinitive as to the relationship between the word frequency and lenition with the data at hand.

• 말소리와 음성과학
• /
• 제8권1호
• /
• pp.45-51
• /
• 2016

This study examined the deletion of onset consonant in the word-internal structure in spontaneous Seoul Korean speech. It used the dataset of speakers in their 20s extracted from the Korean Corpus of Spontaneous Speech (Yun et al., 2015). The proportion of deletion of word-internal onset consonants was analyzed using the linear mixed-effects regression model. The factors that promoted the deletion of onsets were primarily the types of consonants and their phonetic contexts. The results showed that onset deletion was more likely to occur for a lenis velar stop [k] than the other consonants, and in the phonetic contexts, when the preceding vowel was a low central vowel [a]. Moreover, some speakers tended to more frequently delete onset consonants (e.g., [k] and [n]) than other speakers, which reflected individual differences. This study implies that word-internal onsets undergo a process of gradient reduction within individuals' articulatory strategies.

• 말소리와 음성과학
• /
• 제10권4호
• /
• pp.77-89
• /
• 2018

Speech articulators are coordinated for the purpose of segmental constriction in terms of a task. In particular, vertical jaw movements repeatedly contribute to consonantal as well as vocalic constriction. The current study explores vertical jaw movements in conjunction with bilabial constriction in bilabial stop /p/ in the context /a/-to-/a/. Revisiting kinematic data of /p/ collected using the electromagenetic midsagittal articulometer (EMMA) method from seven (four female and three male) speakers of Seoul Korean, we examined maximum vertical jaw position, its relative timing with respect to the upper and lower lips, and lip aperture minima. The results of those dependent variables are recapitulated in terms of linguistic (different word boundaries) and paralinguistic (different speech rates) factors as follows. Firstly, maximum jaw height was lower in the across-word boundary condition (across-word < within-word), but it did not differ as a function of different speech rates (comfortable = fast). Secondly, more reduction in the lip aperture (LA) gesture occurred in fast rate, while word-boundary effects were absent. Thirdly, jaw raising was still in progress after the lips' positional extrema were achieved in the within-word condition, while the former was completed before the latter in the across-word condition. Lastly, relative temporal lags between the jaw and the lips (UL and LL) were more synchronous in fast rate, compared to comfortable rate. When these results are considered together, it is possible to posit that speakers are not tolerant of lenition to the extent that it is potentially realized as a labial approximant in either word-boundary condition while jaw height still manifested lower jaw position in the across-word boundary condition. Early termination of vertical jaw maxima before vertical lower lip maxima across-word condition may be partly responsible for the spatial reduction of jaw raising movements. This may come about as a consequence of an excessive number of factors (e.g., upper lip height (UH), lower lip height (LH), jaw angle (JA)) for the representation of a vector with two degrees of freedom (x, y) engaged in a gesture-based task (e.g., lip aperture (LA)). In the task-dynamic application toolkit, the jaw angle parameter can be assigned numerical values for greater weight in the across-word boundary condition, which in turn gives rise to lower jaw position. Speech rate-dependent spatial reduction in lip aperture may be able to be resolved by means of manipulating activation time of an active tract variable in the gestural score level.

• 지능정보연구
• /
• 제25권4호
• /
• pp.105-122
• /
• 2019

텍스트 데이터가 특정 범주에 속하는지 판별하는 문장 분류에서, 문장의 특징을 어떻게 표현하고 어떤 특징을 선택할 것인가는 분류기의 성능에 많은 영향을 미친다. 특징 선택의 목적은 차원을 축소하여도 데이터를 잘 설명할 수 있는 방안을 찾아내는 것이다. 다양한 방법이 제시되어 왔으며 Fisher Score나 정보 이득(Information Gain) 알고리즘 등을 통해 특징을 선택 하거나 문맥의 의미와 통사론적 정보를 가지는 Word2Vec 모델로 학습된 단어들을 벡터로 표현하여 차원을 축소하는 방안이 활발하게 연구되었다. 사전에 정의된 단어의 긍정 및 부정 점수에 따라 단어의 임베딩을 수정하는 방법 또한 시도하였다. 본 연구는 문장 분류 문제에 대해 선택적 단어 제거를 수행하고 임베딩을 적용하여 문장 분류 정확도를 향상시키는 방안을 제안한다. 텍스트 데이터에서 정보 이득 값이 낮은 단어들을 제거하고 단어 임베딩을 적용하는 방식과, 정보이득 값이 낮은 단어와 코사인 유사도가 높은 주변 단어를 추가로 선택하여 텍스트 데이터에서 제거하고 단어 임베딩을 재구성하는 방식이다. 본 연구에서 제안하는 방안을 수행함에 있어 데이터는 Amazon.com의 'Kindle' 제품에 대한 고객리뷰, IMDB의 영화리뷰, Yelp의 사용자 리뷰를 사용하였다. Amazon.com의 리뷰 데이터는 유용한 득표수가 5개 이상을 만족하고, 전체 득표 중 유용한 득표의 비율이 70% 이상인 리뷰에 대해 유용한 리뷰라고 판단하였다. Yelp의 경우는 유용한 득표수가 5개 이상인 리뷰 약 75만개 중 10만개를 무작위 추출하였다. 학습에 사용한 딥러닝 모델은 CNN, Attention-Based Bidirectional LSTM을 사용하였고, 단어 임베딩은 Word2Vec과 GloVe를 사용하였다. 단어 제거를 수행하지 않고 Word2Vec 및 GloVe 임베딩을 적용한 경우와 본 연구에서 제안하는 선택적으로 단어 제거를 수행하고 Word2Vec 임베딩을 적용한 경우를 비교하여 통계적 유의성을 검정하였다.

• 한국항해항만학회지
• /
• 제41권3호
• /
• pp.109-118
• /
• 2017

주제어(key word, KW)는 해양사고의 주요한 원인을 간단하게 표현하기 위한 단어들의 집합으로 해양안전심판원의 심판관들이 작성한다. KW는 심판관들의 서로 다른 주관적인 견해 때문에 일관성 유지가 어렵고, KW의 수가 너무 많은 문제점이 있다. 이러한 문제를 해결하기 위해서는 최적화된 최소의 공통단어(common word, CW)를 이용한 체계적인 KW 구축 프레임이 필요하다. 본 연구의 목적은 체계적인 KW 구축 프레임 개발에 필요한 CW을 도출하는데 있다. 이러한 목적을 달성하기 위하여 본 연구에서는 파레토(Pareto) 분포함수와 파레토 지수를 이용한 최적의 최소 CW 도출방법을 제안하였다. 총 2,642개의 KW을 수집한 후, 수집한 KW의 세부 단어와 이들의 빈도를 갖는 데이터 세트에서 총 56개의 특징적인 CW를 식별하였다. 56개의 특징적인 CW를 이용한 단어 축소실험을 통해서 평균 58.5%의 축소율을 획득하였고, 축소율에 따라서 추정한 CW는 파레토 차트로 검증하였다. 이를 통해서 체계적인 KW 구축 프레임 개발이 가능할 것으로 기대된다.

• 음성과학
• /
• 제14권4호
• /
• pp.115-128
• /
• 2007

The purposes of this paper were to develop an automatic script to collect such acoustic parameters as duration, intensity, pitch and the first two formant values of English words produced by two native Canadian speakers either alone or in a two-word phrase at a normal speed and to compare those values by the position in the phrases. A Praat script was proposed to obtain the comparable parameters at evenly divided time point of the target word. Results showed that the total duration of the word in the phrase was shorter than that of the word produced alone. That was attributed to the pronunciation style of the native speakers generally placing the primary word stress in the first word position. Also, the reduction ratio of the male speaker depended on the word position in the phrase while the female speaker didn't. Moreover, there were different contours of intensity and pitch by the position of the target word in the phrase while almost the same formant patterns were observed. Further studies would be desirable to examine those parameters of the words in the authentic speech materials.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 추계종합학술대회 논문집
• /
• pp.1019-1022
• /
• 1999

In this paper, we analysis the effect of a dictionary in a handwritten Hangul word recognition problem in terms of its size and the length of the words in it. With our experimental results, we can account for the word recognition rate depending not only on character recognition performance, but also much on the amount of the information that the dictionary contains, as well as the reduction rate of a dictionary.

• 음성과학
• /
• 제7권2호
• /
• pp.7-18
• /
• 2000

The purpose of this study was to investigate the phonological error patterns of korean children with and without specific phonological disorders(SPD). In this study, 29 normally developing children and 10 SPD children were involved. The children were matched the percentage of consonants correct(PCC). 22 picture cards were used to elicit korean consonants in word initial syllable initial, word medial syllable initial, word medial syllable final, word final syllable final positions. The findings were as follows. First, the phonological error patterns of SPD were 1) similar to those of normal children with the same PCC, 2) similar to those of normal children with the lower PCC, or 3) unusual to those of normal children. Second,. korean children showed phonological processes reflecting the korean phonological characteristics: tensification, reduction of the word medial syllable final consonant. This study suggests that both the PCC and error patterns should be considered in assessing phonological abilities of children.