• The Journal of the Acoustical Society of Korea
• /
• v.21 no.3
• /
• pp.330-338
• /
• 2002

Spontaneous speech is ungrammatical as well as serious phonological variations, which make recognition extremely difficult, compared with read speech. In this paper, for conversational speech recognition, we analyze the transcriptions of the real conversational speech, and then classify the characteristics of conversational speech in the speech recognition aspect. Reflecting these features, we obtain the baseline system for conversational speech recognition. The classification consists of long duration of silence, disfluencies and phonological variations; each of them is classified with similar features. To deal with these characteristics, first, we update silence model and append a filled pause model, a garbage model; second, we append multiple phonetic transcriptions to lexicon for most frequent phonological variations. In our experiments, our baseline morpheme error rate (WER) is 31.65%; we obtain MER reductions such as 2.08% for silence and garbage model, 0.73% for filled pause model, and 0.73% for phonological variations. Finally, we obtain 27.92% MER for conversational speech recognition, which will be used as a baseline for further study.

• The KIPS Transactions:PartB
• /
• v.13B no.1 s.104
• /
• pp.63-70
• /
• 2006

In general, a corpus contains lots of linguistic information and is widely used in the field of natural language processing and computational linguistics. The creation of such the corpus, however, is an expensive, labor-intensive and time-consuming work. To alleviate this problem, annotation tools to build corpora with much linguistic information is indispensable. In this paper, we design and implement an annotation tool for establishing a Korean dependency tree-tagged corpus. The most ideal way is to fully automatically create the corpus without annotators' interventions, but as a matter of fact, it is impossible. The proposed tool is semi-automatic like most other annotation tools and is designed to edit errors, which are generated by basic analyzers like part-of-speech tagger and (partial) parser. We also design it to avoid repetitive works while editing the errors and to use it easily and friendly. Using the proposed annotation tool, 10,000 Korean sentences containing over 20 words are annotated with dependency structures. For 2 months, eight annotators have worked every 4 hours a day. We are confident that we can have accurate and consistent annotations as well as reduced labor and time.

• Annual Conference on Human and Language Technology
• /
• 1999.10e
• /
• pp.7-12
• /
• 1999

정보 검색 시스템은 사용자 질의의 키워드들과 문서들의 유사성(similarity)을 기준으로 관련 문서들을 순서화하여 사용자에게 제공한다. 그렇지만 인터넷 검색에 사용되는 질의는 일반적으로 짧기 때문에 보다 유용한 질의를 만들고자 하는 노력이 지금까지 계속되고 있다. 그러나 키워드에 포함된 정보가 제한적이기 때문에 이에 대한 보완책으로 사용자의 적합성 피드백을 이용하는 방법을 널리 사용하고 있다. 본 논문에서는 일반적인 적합성 피드백의 가장 큰 단점인 빈번한 사용자 참여는 지양하고, 시스템에 기반한 적합성 피드백에서 배제한 사용자 참여를 유도하는 검색 문서의 분류 정보에 기반한 용어 클러스터 질의 확장 모델(Term Cluster Query Expansion Model)을 제안한다. 이 방법은 검색 시스템에 의해 검색된 상위 n개의 문서에 대하여 분류기를 이용하여 각각의 문서에 분류 정보를 부여하고, 문서에 부여된 분류 정보를 이용하여 분류 정보의 수(m)만큼으로 문서들을 그룹을 짓는다. 적합성 피드백 알고리즘을 이용하여 m개의 그룹으로부터 각각의 용어 클러스터(Term Cluster)를 생성한다. 이 클러스터가 사용자에게 문서 대신에 피드백의 자료로 제공된다. 실험 결과, 적합성 알고리즘 중 Rocchio방법을 이용할 때 초기 질의보다 나은 성능을 보였지만, 다른 연구에서 보여준 성능 향상은 나타내지 못했다. 그 이유는 분류기의 오류와 문서의 특성상 한 영역으로 규정짓기 어려운 문서가 존재하기 때문이다. 그러나 검색하고자 하는 사용자의 관심 분야나 찾고자 하는 성향이 다르더라도 시스템에 종속되지 않고 유연하게 대처하며 검색 성능(retrieval effectiveness)을 향상시킬 수 있다.사용되고 있어 적응에 문제점을 가지기도 하였다. 본 연구에서는 그 동안 계속되어 온 한글과 한잔의 사용에 관한 논쟁을 언어심리학적인 연구 방법을 통해 조사하였다. 즉, 글을 읽는 속도, 글의 의미를 얼마나 정확하게 이해했는지, 어느 것이 더 기억에 오래 남는지를 측정하여 어느 쪽의 입장이 옮은 지를 판단하는 것이다. 실험 결과는 문장을 읽는 시간에서는 한글 전용문인 경우에 월등히 빨랐다. 그러나. 내용에 대한 기억 검사에서는 국한 혼용 조건에서 더 우수하였다. 반면에, 이해력 검사에서는 천장 효과(Ceiling effect)로 두 조건간에 차이가 없었다. 따라서, 본 실험 결과에 따르면, 글의 읽기 속도가 중요한 문서에서는 한글 전용이 좋은 반면에 글의 내용 기억이 강조되는 경우에는 한자를 혼용하는 것이 더 효율적이다.이 높은 활성을 보였다. 7. 이상을 종합하여 볼 때 고구마 끝순에는 페놀화합물이 다량 함유되어 있어 높은 항산화 활성을 가지며, 아질산염소거능 및 ACE저해활성과 같은 생리적 효과도 높아 기능성 채소로 이용하기에 충분한 가치가 있다고 판단된다.등의 관련 질환의 예방, 치료용 의약품 개발과 기능성 식품에 효과적으로 이용될 수 있음을 시사한다.tall fescue 23%, Kentucky bluegrass 6%, perennial ryegrass 8%) 및 white clover 23%를 유지하였다. 이상의 결과를 종합할 때, 초종과 파종비율에 따른 혼파초지의 건물수량과 사료가치의 차이를 확인할 수 있었으며, 레드 클로버 + 혼파 초지가 건물수량과 사료가치를 높이는데 효과적이었다.\ell}$ 이었으며 , yeast extract 첨가(添加)하여 배양시(培養時)는 yeast extract

• The Journal of Korean society of community based occupational therapy
• /
• v.8 no.1
• /
• pp.45-54
• /
• 2018

Objective : It is one of the most important job tasks to write to occupational therapist, so I want to apply the writing program to the students who use mobile language as the main communication and to investigate the effect. Methods : This study was conducted with 7 freshman students for a total of 10 sessions, once a week, for 2 hours at a time. In addition, after reading the recommended books, I made a total of two manuscripts with the book report and then carried out the supplementary instruction. Changes in the writing program were made using self - questionnaires and changes in the writing of the manuscripts were confirmed by the number of times. Results : As a result of the self-questionnaire, the participants considered the logical aspect of the writing and the consistency of the writing after participating in the writing program, and after the writing, the grammatical aspect was reviewed and the sentence was revised. In addition, the number of secondary corrections was reduced by an average of 7 times more than the number of primary corrections. Conclusion : In order to create a document which is one of the important tasks for occupational therapist, systematic education will be needed to create a more logical and grammatical error-free article.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.115-118
• /
• 2002

Parameters used in a speaker recognition system are desirable expressing speaker's characteristics filly and have in a speech. That is to say, if inter-speaker than intra-speaker variance a big characteristic, it is useful to distinguish between speakers. Also, to make minimum error between speakers, it is required the improved recognition technology as well as the distinguishing characteristics. When we see the result of recent simulation performance, we obtain more exact performance by using dynamic characteristics and constant characteristics by a speaking habit. Therefore we suggest it to solve this problem as followings. The prosodic information is used by a characteristic vector of speech. Characteristics vector generally using in speaker recognition system is a modeling spectrum information and is working for a high performance in non-noise circumstance. However, it is found a problem that characteristic vector is distorted in noise circumstance and it makes a reduction of recognition rate. In this paper, we change pitch line divided by segment which can estimate a dynamic characteristic and it is used as a recognition characteristic. we confirmed that the dynamic characteristic is very robust in noise circumstance with a simulation. We make a decision of acceptance or rejection by comparing test pattern and recognition rate using the proposed algorithm has more improvement than using spectrum and prosodic information. Especially stational recognition rate can be obtained in noise circumstance through the simulation.

• School Mathematics
• /
• v.18 no.2
• /
• pp.257-275
• /
• 2016

The purpose of this study was to investigate the impact of multiple representations on students' understanding of fractions, decimals, and percent. The instructional approach integrating multiple representations was compared to traditional algorithmic instruction, a form of direct instruction. To examine and compare the impact of multiple representations instruction with traditional algorithmic instruction, pre and post tests consisting of five similar items were administered with 87 middle school students. Students' scores in these two tests and their problem solving processes were analyzed quantitatively and qualitatively. The quantitative results indicated that students taught by traditional algorithmic instruction showed higher scores on the post-test than students in the multiple representations group. Furthermore, findings suggest that instruction using multiple representations does not guarantee a positive impact on students' understanding of mathematical concepts. Qualitative results suggest that the limited use of multiple representations during a class may have hindered students from applying their use in novel problem situations. Therefore, when using multiple representations, teachers should employ more diverse examples and practice with multiple representations to help students to use them without error.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.30 no.1
• /
• pp.5-27
• /
• 2019

This study analyzed the effects of comic book reading program on Korean proficiency and acculturation of youth with immigration background, by conducting a six-month reading program with five teenagers with immigration background. Ten comic books were selected from published by School Library Journal, based on the themes - that are related to the lives of youth with immigration background - and interests of participating teens. According to the literacy skills test conducted before and after the reading program, the participating teens' Korean proficiency has generally improved, particularly in the areas of interpretation and vocabulary. In terms of writing, grammatically incorrect sentences, phrases, and expressions have declined. Most participants showed stable adjustment to Korean culture, but one participant who felt still insecure of her ethnic identity deeply empathized with one of the characters of the books, and shared the difficulties of living as an outsider of a society. The participants of this research learned or rediscovered the joy of reading through this comic book reading program; at the end of the program, many of them expanded their interest in reading novels, books without any illustrations.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.29 no.7
• /
• pp.843-856
• /
• 2023

The global surge in maritime traffic has resulted in an increased number of ship collisions, leading to significant economic, environmental, physical, and human damage. The causes of these maritime accidents are multifaceted, often arising from a combination of crew judgment errors, negligence, complexity of navigation routes, weather conditions, and technical deficiencies in the vessels. Given the intricate nuances and contextual information inherent in each incident, a methodology capable of deeply understanding the semantics and context of sentences is imperative. Accordingly, this study utilized the SentenceBERT model to analyze maritime safety tribunal decisions over the last 20 years in the Busan Sea area, which encapsulated data on ship collision incidents. The analysis revealed important keywords potentially responsible for these incidents. Cluster analysis based on the frequency of specific keyword appearances was conducted and visualized. This information can serve as foundational data for the preemptive identification of accident causes and the development of strategies for collision prevention and response.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.2
• /
• pp.155-163
• /
• 2009

In text-to-speech systems, the conversion of text into prosodic parameters is necessarily composed of three steps. These are the placement of prosodic boundaries. the determination of segmental durations, and the specification of fundamental frequency contours. Prosodic boundaries. as the most important and basic parameter. affect the estimation of durations and fundamental frequency. Break prediction is an important step in text-to-speech systems as break indices (BIs) have a great influence on how to correctly represent prosodic phrase boundaries, However. an accurate prediction is difficult since BIs are often chosen according to the meaning of a sentence or the reading style of the speaker. In Japanese, the prediction of an accentual phrase boundary (APB) and major phrase boundary (MPB) is particularly difficult. Thus, this paper presents a method to complement the prediction errors of an APB and MPB. First, we define a subtle BI in which it is difficult to decide between an APB and MPB clearly as a variable break (VB), and an explicit BI as a fixed break (FB). The VB is chosen using the classification and regression tree, and multiple prosodic targets in relation to the pith and duration are then generated. Finally. unit-selection is conducted using multiple prosodic targets. In the MOS test result. the original speech scored a 4,99. while proposed method scored a 4.25 and conventional method scored a 4.01. The experimental results show that the proposed method improves the naturalness of synthesized speech.

• Journal of Korean Ophthalmic Optics Society
• /
• v.18 no.2
• /
• pp.125-136
• /
• 2013

Purpose: The aim of this study was to investigate the effect of improving read speed with color filter or without color filter to improve reading disorder of teenager who were diagnosed as Meares-Irlen syndrome through survey inspection with Meares-Irlen syndrome visual stress (MISViS) score. Methods: MISViS subjects were selected from screening survey MISViS results given above 2.13 in the clinical criteria scores (MISViS score). Reading speed were measured quickly and efficiently the rate of reading via test in which randomly ordered common words are read aloud during a minute. Each of the subjects were worn a filter of the lowest concentration in each color filter group composed of 15 groups. Results: MISViS score of MISViS group and control group were 2.57 and 0.66, respectively. Results of reading speed with filter and without filter in MISViS group were $102.27{\pm}27.86$ wpm and $118.87{\pm}26.99$ wpm (p=0.001), respectively, as well as were $132.93{\pm}6.88$ wpm and $133.43{\pm}6.64$ wpm (p=0.131) in the normal group. Associated with error changes with filter and without filter between two groups, skipping in MISViS Group were from $0.25{\pm}0.62$ times to 0 times (p=0.191), Errors were from $1.83{\pm}1.69$ times to $0.17{\pm}0.38$ times (p = 0.004) and, repetitions were 0. skipping in control group were 0 times, errors were from $0.21{\pm}0.43$ times to $0.07{\pm}0.27$ times (p=0.336) and, repetitions were from $0.14{\pm}0.36$ times to 0 (p=0.165). The filter of blue series chosen in MISViS group had higher percentage (40%), whereas, subjects in normal group were more likely to prefer the filter of gray color (29%). Conclusions: This study showed that MISViS score have been used as a significant diagnosis for Irlen syndrome screening. This study found that wearing suitable color filter for MISViS patients were useful to improve learning with regard to reading. Unique color filter selection for MISViS subjects must be carefully considered since fit color filter are different personally.