• The KIPS Transactions:PartB
• /
• v.16B no.5
• /
• pp.427-434
• /
• 2009

With the rapid evolution of the Internet and mobile environments, text including spelling errors such as newly-coined words and abbreviated words are widely used. These spelling errors make it difficult to develop NLP (natural language processing) applications because they decrease the readability of texts. To resolve this problem, we propose a spelling error correction model using a spelling error correction dictionary and a newspaper corpus. The proposed model has the advantage that the cost of data construction are not high because it uses a newspaper corpus, which we can easily obtain, as a training corpus. In addition, the proposed model has an advantage that additional external modules such as a morphological analyzer and a word-spacing error correction system are not required because it uses a simple string matching method based on a correction dictionary. In the experiments with a newspaper corpus and a short message corpus collected from real mobile phones, the proposed model has been shown good performances (a miss-correction rate of 7.3%, a F1-measure of 97.3%, and a false positive rate of 1.1%) in the various evaluation measures.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.3
• /
• pp.357-362
• /
• 2004

Language models are essential in predicting the next word in a spoken sentence, thereby enhancing the speech recognition accuracy, among other things. However, spoken language domains are too numerous, and therefore developers suffer from the lack of corpora with sufficient sizes. This paper proposes a method of combining two n-gram language models, one constructed from a very small corpus of the right domain of interest, the other constructed from a large but less adequate corpus, resulting in a significantly enhanced language model. This method is based on the observation that a small corpus from the right domain has high quality n-grams but has serious sparseness problem, while a large corpus from a different domain has more n-gram statistics but incorrectly biased. With our approach, two n-gram statistics are combined by extending the idea of Katz's backoff and therefore is called a dual-source backoff. We ran experiments with 3-gram language models constructed from newspaper corpora of several million to tens of million words together with models from smaller broadcast news corpora. The target domain was broadcast news. We obtained significant improvement (30%) by incorporating a small corpus around one thirtieth size of the newspaper corpus.

• Journal of KIISE
• /
• v.43 no.1
• /
• pp.71-79
• /
• 2016

Disambiguation of homographs is an important job in Korean semantic processing and has been researched for long time. Recently, machine learning approaches have demonstrated good results in accuracy and speed. Other knowledge-based approaches are being researched for untrained words. This paper proposes a hybrid method based on the machine learning approach that uses a lexical semantic network. The use of a hybrid approach creates an additional corpus from subcategorization information and trains this additional corpus. A homograph tagging phase uses the hypernym of the homograph and an additional corpus. Experimentation with the Sejong Corpus and UWordMap demonstrates the hybrid method is to be effective with an increase in accuracy from 96.51% to 96.52%.

• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.1-7
• /
• 2021

In the latest trend of machine translation research, the model is pretrained through a large mono lingual corpus and then finetuned with a parallel corpus. Although many studies tend to increase the amount of data used in the pretraining stage, it is hard to say that the amount of data must be increased to improve machine translation performance. In this study, through an experiment based on the mBART model using parallel corpus filtering, we propose that high quality data can yield better machine translation performance, even utilizing smaller amount of data. We propose that it is important to consider the quality of data rather than the amount of data, and it can be used as a guideline for building a training corpus.

• 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.

• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.23-29
• /
• 2021

Recent deep learning-based natural language processing studies are conducting research to improve performance by training large amounts of data from various sources together. However, there is a possibility that the methodology of learning by combining data from various sources into one may prevent performance improvement. In the case of machine translation, data deviation occurs due to differences in translation(liberal, literal), style(colloquial, written, formal, etc.), domains, etc. Combining these corpora into one for learning can adversely affect performance. In this paper, we propose a new Corpus Weight Balance(CWB) method that considers the balance between parallel corpora in machine translation. As a result of the experiment, the model trained with balanced corpus showed better performance than the existing model. In addition, we propose an additional corpus construction process that enables coexistence with the human translation market, which can build high-quality parallel corpus even with a monolingual corpus.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.12
• /
• pp.1254-1258
• /
• 2010

In information retrieval, different surface forms of the same object can cause poor performance of systems. In this paper, we propose the method extracting alternative words using translation words as features of each word extracted from parallel corpus, korean/english title pair of patent information. Also, we propose an association word filtering method to remove association words from an alternative word list. Evaluation results show that the proposed method outperforms other alternative word extraction methods.

• Annual Conference on Human and Language Technology
• /
• 1999.10d
• /
• pp.117-122
• /
• 1999

본 연구는 형태소 분석에 필요한 언어 지식과 품사 태깅에 필요한 확률 정보를 별도의 언어 지식 추가 없이 학습 말뭉치를 통해서 얻어내는 방법을 제안한다. 먼저 품사 부착된 학습 말뭉치로부터 형태소 사전과 결합 정보를 추출한다. 그리고 자주 발생하는 어절 및 해석상 모호성이 많은 어절에 대해서는 학습 말뭉치에서 발견된 형태소 분석 결과를 저장하여 형태소 분석에 소요되는 시간과 형태소 분석의 정확률을 높인다. 또한 미등록어의 많은 부분을 차지하는 인명, 지명, 조직명에 대해서는 정보 추출 분야에서 사용하는 고유 명사 분류법으로 해결한다. 품사 태깅을 위해서는 품사열 정보와 품사열 정보로는 해결할 수 없는 경우를 위한 어휘 정보를 학습 말뭉치에서 추출한다. 품사열 정보와 어휘 정보는 정형화 과정을 거쳐 최대 엔트로피 모델의 자질로 사용되어 품사 태깅 시스템을 위한 확률 분포를 구성한다. 본 연구에서 제안하는 방법은 학습 말뭉치를 기반으로 한다는 특성에 의해 다양한 영역에 사용하기 쉽다. 또한 어휘 정보로 품사 문맥 정보를 보완하기 때문에 품사 분류 체계와 형태소 해석 규칙에 영향을 적게 받는다는 장점을 가진다. MATEC '99 데이터 실험 결과 형태소 단위로 94%의 재현률과 93%의 정확률을 얻을 수 있었다.

• Annual Conference on Human and Language Technology
• /
• 2012.10a
• /
• pp.35-39
• /
• 2012

의미역 결정에 있어 의미 표지 부착 말뭉치는 필수적이지만 한국어 의미 표지 부착 말뭉치는 영어나 중국어와 같은 언어에 비하여 구축이 미비한 상황이다. 본 논문에서는 한국어 의미 분석을 위한 한국어 Proposition Bank(이하 PropBank)와 세종 의미 표지 부착 말뭉치의 구축을 위한 소프트웨어 도구를 개발하였다. 본 논문에서 구현한 도구는 문장 성분의 의존관계를 이용하여 주어진 술어에 대한 논항을 찾아주고, PropBank 프레임 파일과 세종 용언 격틀 사전을 활용하여 사용자가 능률적으로 한국어 PropBank와 세종 의미 표지 부착 말뭉치를 구축할 수 있도록 하였다.

• Annual Conference on Human and Language Technology
• /
• 2005.10a
• /
• pp.151-158
• /
• 2005

본 논문에서는 2단계 구문분석을 통한 구문분석 말뭉치 구축도구를 제안한다. 제안하는 방법은 대량의 구문분석 말뭉치를 수동으로 구축할 때 요구되는 작성자의 수작업을 줄이는 것을 목적으로 한다. 도구는 입력 문장을 문장 분할기준에 따라 분할하는 문장 분할 단계, 각 부분에 대해 자동 구문분석을 수행하는 부분 구문구조 생성 단계, 각 부분 구문구조를 통합하여 완전한 구문구조를 얻는 부분 통합 단계로 이루어져 있다. 자동 구문분석은 자질기반 한국어 구문분석모델을 이용하였고 문장을 부분으로 분할할 때는 문장 분할기준을 말뭉치에서 자동추출 하고 간단한 검증을 거쳐 적용하는 방법을 택하였다. 구문분석 말뭉치 구축의 각 단계에서 자동 구문 분석기가 출력한 결과를 작성자가 취소, 재구축 가능하게 하였다.