• 한국정보과학회 언어공학연구회:학술대회논문집(한글 및 한국어 정보처리)
• /
• 한국정보과학회언어공학연구회 2017년도 제29회 한글 및 한국어 정보처리 학술대회
• /
• pp.103-107
• /
• 2017

한국어 형태소 분석 및 구구조 구문 분석은 한국어 자연어처리에서 난이도가 높은 작업들로서 최근에는 해당 문제들을 출력열 생성 문제로 바꾸어 sequence-to-sequence 모델을 이용한 end-to-end 방식의 접근법들이 연구되었다. 한국어 형태소 분석 및 구구조 구문 분석을 출력열 생성 문제로 바꿀 시 해당 출력 결과는 하나의 열로서 합쳐질 수가 있다. 본 논문에서는 sequence-to-sequence 모델을 이용하여 한국어 형태소 분석 및 구구조 구문 분석을 동시에 처리하는 모델을 제안한다. 실험 결과 한국어 형태소 분석과 구구조 구문 분석을 동시에 처리할 시 형태소 분석이 구구조 구문 분석에 영향을 주는 것을 확인 하였으며, 구구조 구문 분석 또한 형태소 분석에 영향을 주어 서로 영향을 줄 수 있음을 확인하였다.

• The Journal of the Acoustical Society of Korea
• /
• 제16권1E호
• /
• pp.57-63
• /
• 1997

In this paper, we investigated performance of a vocabulary-independent speech recognizer with speaker adaptation. The vocabulary-independent speech recognizer does not require task-oriented speech databases to estimate HMM parameters, but adapts the parameters recursively by using input speech and recognition results. The recognizer has the advantage that it relieves efforts to record the speech databases and can be easily adapted to a new task and a new speaker with different recognition vocabulary without losing recognition accuracies. Experimental results showed that the vocabulary-independent speech recognizer with supervised offline speaker adaptation reduced 40% of recognition errors when 80 words from the same vocabulary as test data were used as adaptation data. The recognizer with unsupervised online speaker adaptation reduced abut 43% of recognition errors. This performance is comparable to that of a speaker-independent speech recognizer trained by a task-oriented speech database.

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2000년도 춘계학술대회논문집 - 한국공작기계학회
• /
• pp.543-548
• /
• 2000

In this study, an interactive programming method for robot in electronic part assembly task is proposed. Many of industrial robots are still taught and programmed by a teach pendant. The robot is guided by a human operator to the desired application locations. These motions are recorded and are later edited, within the robotic language using in the robot controller, and play back repetitively to perform robot task. This conventional teaching method is time-consuming and somewhat dangerous. In the proposed method, the operator teaches the desired locations on the image acquired through CCD camera mounted on the robot hand. The robotic language program is automatically generated and downloaded to the robot controller. This teaching process is implemented through an off-line programming software. The OLP is developed for an robotic assembly system used in this study. In order to transform the location on image coordinates into robot coordinates, a calibration process is established. The proposed teaching method is implemented and evaluated on an assembly system for soldering electronic parts on a circuit board. A six-axis articulated robot executes assembly task according to the off-line teaching in the system.

• 대한전기학회논문지
• /
• 제38권9호
• /
• pp.740-752
• /
• 1989

Many works have been reported in various fields on the subject of controlling a robot with high-level robot languages. This paper presents one such effort and explains the development of an automatic robot programming system which utilizes the concept of the task level language. This system is expected to act as an intelligent supporting tool in robot programming and be put into practical use. Emphasis is placed on the role of the programming system as a tool that generates the executable robot program according to the user specified tasks. Several task level commands are used in the developed system, and the resulting inflexibility is complemented by the motion level commands of the motion level robot languages. Thus, the advantages of both task and motion level languages are utilized, and no knowledge of specific language grammer is needed even when using motion level commands. To increase the usability of the developed system, various methods are provided for supplementing the programming system using taught data.

• ETRI Journal
• /
• 제44권5호
• /
• pp.794-804
• /
• 2022

Transformer models have had a great impact on natural language processing (NLP) in recent years by realizing outstanding and efficient contextualized language models. Recent studies have used transformer-based language models for various NLP tasks, including Persian named entity recognition (NER). However, in complex tasks, for example, NER, it is difficult to determine which contextualized embedding will produce the best representation for the tasks. Considering the lack of comparative studies to investigate the use of different contextualized pretrained models with sequence modeling classifiers, we conducted a comparative study about using different classifiers and embedding models. In this paper, we use different transformer-based language models tuned with different classifiers, and we evaluate these models on the Persian NER task. We perform a comparative analysis to assess the impact of text representation and text classification methods on Persian NER performance. We train and evaluate the models on three different Persian NER datasets, that is, MoNa, Peyma, and Arman. Experimental results demonstrate that XLM-R with a linear layer and conditional random field (CRF) layer exhibited the best performance. This model achieved phrase-based F-measures of 70.04, 86.37, and 79.25 and word-based F scores of 78, 84.02, and 89.73 on the MoNa, Peyma, and Arman datasets, respectively. These results represent state-of-the-art performance on the Persian NER task.

• 한국콘텐츠학회논문지
• /
• 제16권5호
• /
• pp.58-67
• /
• 2016

후천성 뇌손상(ABI) 환자의 인지-언어 능력은 하위 영역 간 상호작용의 맥락에서 고려되어야 하는데, 특히 화용언어는 통합적 인지 처리를 관장하는 집행기능과 직결된다. 본 연구에서는 ABI 환자를 대상으로 화용언어와 집행기능 간의 상관성을 규명하고자 하였다. 이를 위해 만 55세 이상의 ABI 환자군 35명(뇌졸중에 의한 실어증 21명, TBI 14명)을 대상으로 화용언어와 집행기능 관련 5개 하위 과제를 평가하였다. 그 결과, 실어증 집단은 비유언어 이해 및 기능/상징언어가 실행화 과제와 유의한 상관성을 보였고, TBI 집단은 모든 과제 간에 유의한 상관성이 있었다. 실어증 집단의 비유언어 이해는 실행화, 그리고 TBI 집단의 비유언어 표현과 기능/상징언어는 각각 계획화와 실행화를 가장 잘 예측하는 과제였다. 본 연구를 통해 ABI 이후의 인지-언어적 중재 시 상호 보완적으로 활용할 수 있는 화용언어 및 집행기능 과제들을 제시할 수 있었다.

• International Journal of Advanced Culture Technology
• /
• 제8권2호
• /
• pp.101-108
• /
• 2020

One difficulty in language learning is the training of reading ability. The improvement on this ability directly affects the process and effect of language learning. At the same time, there are numerous difficulties in actual learning and teaching. Depending on current research, there is two ideas that can utilize to enhance the reading efficiency of learners. One is to amend objective factors; the other is to change subjective factors. Compared with the two ideas, idiosyncratic factors are more manipulable and controllable, so it is more valuable to conduct researches on this. But among the many subjective factors, the degree of their effectiveness is not the same, so this article attempts to compare and analyze the driving effects of two important subjective factors (peer-driven and task-driven) on reading performance. The results show that both factors can have a positive impact on reading comprehension, but different in driving effects. The task-driven has obvious short-term effectiveness; while peer-driven needs to establish its long-term effect on the basis of early coordination and cooperation among team members. Therefore, in order to maximize the achievement of learning, it is necessary to combine strengths and avoid weaknesses according to the characteristics of two factors, so as to help learners improve reading ability most efficiently.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제12권8호
• /
• pp.341-354
• /
• 2023

문서의 제목은 문서의 내용을 가장 효율적으로 요약하여 제공해 준다. 이때 독자들이 선호하는 스타일과 언어에 따라 문서의 제목을 다르게 제공해 준다면, 독자들은 문서의 내용을 좀 더 쉽게 예측할 수 있다. 본 연구에서는 문서가 주어졌을 때 언어와 스타일에 따라 제목을 자동 생성하는'교차 언어 및 스타일 기반의 제목 생성 모델을 제안한다. 모델을 학습하기 위해서는 같은 내용을 다른 언어와 다른 스타일로 작성한 병렬데이터가 필요하다. 그러나 이러한 종류의 병렬데이터는 구축하기 매우 어렵다. 반면, 단일 언어와 단일 스타일로 구축된 제목 생성 데이터는 많으므로 본 연구에서는 제로샷(zero-shot) 학습으로 제목 생성을 수행하고자 한다. 교차 언어 및 스타일 기반의 제목 생성을 학습하기 위해 다중 언어로 사전 학습된 트랜스포머 모델에 각 언어, 스타일, 기계번역을 위한 어댑터를 추가하였다. 기계 번역용 병렬데이터를 이용하여 기계번역을 먼저 학습한 후, 동일 스타일의 제목 생성을 학습하였다. 이때, 필요한 어댑터만을 학습하고 다른 부분의 파라미터는 모두 고정시킨다. 교차 언어 및 스타일 기반의 제목을 생성할 때에는 목적 언어와 목적 스타일에 해당하는 어댑터만을 활성화시킨다. 실험 결과로는 각 모델을 따로 학습시켜 파이프라인으로 연결시킨 베이스라인에 비해 본 연구에서 제안한 제로샷 제목 생성의 성능이 크게 떨어지지 않았다. 최근 대규모 언어 모델의 등장으로 인한 자연어 생성에서의 많은 변화가 있다. 그러나 제한된 자원과 제한된 데이터만을 이용하여 자연어 생성의 성능을 개선하는 연구는 계속되어야 하며, 그런 점에서 본 연구의 의의를 모색한다.

• ETRI Journal
• /
• 제42권1호
• /
• pp.90-100
• /
• 2020

Named entity recognition (NER) continues to be an important task in natural language processing because it is featured as a subtask and/or subproblem in information extraction and machine translation. In Urdu language processing, it is a very difficult task. This paper proposes various deep recurrent neural network (DRNN) learning models with word embedding. Experimental results demonstrate that they improve upon current state-of-the-art NER approaches for Urdu. The DRRN models evaluated include forward and bidirectional extensions of the long short-term memory and back propagation through time approaches. The proposed models consider both language-dependent features, such as part-of-speech tags, and language-independent features, such as the "context windows" of words. The effectiveness of the DRNN models with word embedding for NER in Urdu is demonstrated using three datasets. The results reveal that the proposed approach significantly outperforms previous conditional random field and artificial neural network approaches. The best f-measure values achieved on the three benchmark datasets using the proposed deep learning approaches are 81.1%, 79.94%, and 63.21%, respectively.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2007년도 한국음성과학회 공동학술대회 발표논문집
• /
• pp.78-83
• /
• 2007

The purpose of present study was to investigate the characteristics of disfluency between the Korean-English bilingual and Korean monolingual children, matched by their chronological age with the bilingual children. Twenty-eight children, 14 bilingual children and 14 monolingual children participated in this study. The experimental tasks consisted of the play situation and the task situation. The conclusion is (a) The score of total disfluency of the bilingual was significantly higher than that of the monolingual. The score of normal disfluency of the bilingual was significantly higher than that of the monolingual. The most frequent type is Interjection in both groups. All shows higher score in the task situation than the play situation. The bilingual children have quantitative and qualitative differences in disfluency score and types from the monolingual. (b) The bilingual were divided into two groups such as 6 Korean-dominant bilingual and 8 English-dominant bilingual. All shows more disfluency in their non-dominant language. The most frequent type is Interjection in both groups. (c) The higher the chronological age and the expressive language test score is, the lower the disfluency score is. The earlier the exposure age to the 2nd language is, the higher the disfluency score is. There is no correlation between resident month at foreign country and the disfluency.