• Korean Journal of Cognitive Science
• /
• v.21 no.1
• /
• pp.157-175
• /
• 2010

Understanding dialogue participant's emotion is important as well as decoding the explicit message in human communication. It is well known that non-verbal elements are more suitable for conveying speaker's emotions than verbal elements. Written texts, however, contain a variety of linguistic units that express emotions. This study aims at analyzing components for constructing an emotion ontology, that provides us with numerous applications in Human Language Technology. A majority of the previous work in text-based emotion processing focused on the classification of emotions, the construction of a dictionary describing emotion, and the retrieval of those lexica in texts through keyword spotting and/or syntactic parsing techniques. The retrieved or computed emotions based on that process did not show good results in terms of accuracy. Thus, more sophisticate components analysis is proposed and the linguistic factors are introduced in this study. (1) 5 linguistic types of emotion expressions are differentiated in terms of target (verbal/non-verbal) and the method (expressive/descriptive/iconic). The correlations among them as well as their correlation with the non-verbal expressive type are also determined. This characteristic is expected to guarantees more adaptability to our ontology in multi-modal environments. (2) As emotion-related components, this study proposes 24 emotion types, the 5-scale intensity (-2~+2), and the 3-scale polarity (positive/negative/neutral) which can describe a variety of emotions in more detail and in standardized way. (3) We introduce verbal expression-related components, such as 'experiencer', 'description target', 'description method' and 'linguistic features', which can classify and tag appropriately verbal expressions of emotions. (4) Adopting the linguistic tag sets proposed by ISO and TEI and providing the mapping table between our classification of emotions and Plutchik's, our ontology can be easily employed for multilingual processing.

• Journal of the Korea Convergence Society
• /
• v.12 no.11
• /
• pp.35-43
• /
• 2021

Recently, there has been a growing interest in zero-shot cross-lingual transfer, which leverages cross-lingual language models (CLLMs) to perform downstream tasks that are not trained in a specific language. In this paper, we point out the limitations of the data-centric aspect of quality estimation (QE), and perform zero-shot cross-lingual transfer even in environments where it is difficult to construct QE data. Few studies have dealt with zero-shots in QE, and after fine-tuning the English-German QE dataset, we perform zero-shot transfer leveraging CLLMs. We conduct comparative analysis between various CLLMs. We also perform zero-shot transfer on language pairs with different sized resources and analyze results based on the linguistic characteristics of each language. Experimental results showed the highest performance in multilingual BART and multillingual BERT, and we induced QE to be performed even when QE learning for a specific language pair was not performed at all.

• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.5
• /
• pp.15-30
• /
• 2023

Recently, generative models based on the Transformer architecture, such as ChatGPT, have been gaining significant attention. The Transformer architecture has been applied to various neural network models, including Google's BERT(Bidirectional Encoder Representations from Transformers) sentence generation model. In this paper, a method is proposed to create a text binary classification model for determining whether a comment on Korean movie review is positive or negative. To accomplish this, a pre-trained multilingual BERT sentence generation model is fine-tuned and transfer learned using a new Korean training dataset. To achieve this, a pre-trained BERT-Base model for multilingual sentence generation with 104 languages, 12 layers, 768 hidden, 12 attention heads, and 110M parameters is used. To change the pre-trained BERT-Base model into a text classification model, the input and output layers were fine-tuned, resulting in the creation of a new model with 178 million parameters. Using the fine-tuned model, with a maximum word count of 128, a batch size of 16, and 5 epochs, transfer learning is conducted with 10,000 training data and 5,000 testing data. A text sentiment binary classification model for Korean movie review with an accuracy of 0.9582, a loss of 0.1177, and an F1 score of 0.81 has been created. As a result of performing transfer learning with a dataset five times larger, a model with an accuracy of 0.9562, a loss of 0.1202, and an F1 score of 0.86 has been generated.

• Annual Conference on Human and Language Technology
• /
• 2017.10a
• /
• pp.143-146
• /
• 2017

심층학습 모델 중 LSTM-CRF는 개체명 인식, 품사 태깅과 같은 sequence labeling에서 우수한 성능을 보이고 있다. 한국어 개체명 인식에 대해서도 LSTM-CRF 모델을 기본 골격으로 단어, 형태소, 자모음, 품사, 기구축 사전 정보 등 다양한 정보와 외부 자원을 활용하여 성능을 높이는 연구가 진행되고 있다. 그러나 이런 방법은 언어 자원과 성능이 좋은 자연어 처리 모듈(형태소 세그먼트, 품사 태거 등)이 없으면 사용할 수 없다. 본 논문에서는 LSTM-CRF와 최소한의 언어 자원을 사용하여 다국어에 대한 개체명 인식에 대한 성능을 평가한다. LSTM-CRF의 입력은 문자 기반의 n-gram 표상으로, 성능 평가에는 unigram 표상과 bigram 표상을 사용했다. 한국어, 일본어, 중국어에 대해 개체명 인식 성능 평가를 한 결과 한국어의 경우 bigram을 사용했을 때 78.54%의 성능을, 일본어와 중국어는 unigram을 사용했을 때 각 63.2%, 26.65%의 성능을 보였다.

• International Journal of Internet, Broadcasting and Communication
• /
• v.9 no.2
• /
• pp.27-36
• /
• 2017

The world's unique geographical, multilingual, multiethnic, multiracial and multi religious Himalayan country Nepal has more than 100 years history on telephony service and it has been formulating appropriate policy and regulation for the adoption of new technology, introducing the competitive market environment for the overall development of Information and Communication Technology (ICT) infrastructures and application of ICT service and tools for socio-economic transformation. The Nepalese market seems to be continuously growing and having huge demand of mobile telephony and internet subscriptions trend. The ICT infrastructure development in difficult geographical area is quite challenging and thus operators are focusing mobile telephony and mobile internet services. Nepal has been doing its best effort on formulating policy and regulation, adoption key strategies for ICT sector development and at the same time joining hands with international and regional bodies such as ITU, SAARC etc for ICT sector development. Due to geographical diversity, policy and regulatory barriers in some extent, power supply constraints and low affordability from customers on ICT tools and services, Nepal has been facing challenges on ICT infrastructure development. However, the national statistics on ICT, Networked Readiness Index and ICT Development Index show that Nepal has done quite good progress and is keeping its pace on ICT development despite the these challenges. Moreover, there seems to be quite uncovered market segments on internet service and big opportunity on ICT sector development in Nepal in the days to come.

• Journal of the Korean Society for information Management
• /
• v.15 no.3
• /
• pp.7-26
• /
• 1998

The large scale documents of the data repository are utilized to the diverse applications. In the cross-language information retrieval, if the words of a query contain polymorphic meanings, the system needs multilingual corpus to exactly translate to the target words. We constructed the financial comparable corpus, called KFCM(Korean Financial Corpus corresponding to MLCC Corpus), comparing to the MLCC Polylingual Documents which consisted with the 6 European languages. It is independently constructed under the DTD of MLCC comparable corpus, and can be utilized to the cross-language information retrieval. In this paper, we discussed about the application and construction procedures of KFCM which is public domain data.

• 한국어정보학회:학술대회논문집
• /
• 2017.10a
• /
• pp.143-146
• /
• 2017

심층학습 모델 중 LSTM-CRF는 개체명 인식, 품사 태깅과 같은 sequence labeling에서 우수한 성능을 보이고 있다. 한국어 개체명 인식에 대해서도 LSTM-CRF 모델을 기본 골격으로 단어, 형태소, 자모음, 품사, 기구축 사전 정보 등 다양한 정보와 외부 자원을 활용하여 성능을 높이는 연구가 진행되고 있다. 그러나 이런 방법은 언어 자원과 성능이 좋은 자연어 처리 모듈(형태소 세그먼트, 품사 태거 등)이 없으면 사용할 수 없다. 본 논문에서는 LSTM-CRF와 최소한의 언어 자원을 사용하여 다국어에 대한 개체명 인식에 대한 성능을 평가한다. LSTM-CRF의 입력은 문자 기반의 n-gram 표상으로, 성능 평가에는 unigram 표상과 bigram 표상을 사용했다. 한국어, 일본어, 중국어에 대해 개체명 인식 성능 평가를 한 결과 한국어의 경우 bigram을 사용했을 때 78.54%의 성능을, 일본어와 중국어는 unigram을 사용했을 때 각 63.2%, 26.65%의 성능을 보였다.

• Journal of the Korean Society for information Management
• /
• v.26 no.3
• /
• pp.25-43
• /
• 2009

The purpose of this study is to suggest how the classic thesaurus structure of terms and links can be mined and updated from Wikipedia encyclopedia, which is the best practice of collective intelligence. In a comparison with ASIS&T thesaurus, it was found that Wikipedia contains a substantial coverage of domain-specific concepts and semantic relations. Furthermore, it was resulted that the structural characteristics of Wikipedia, such as redirects, categories, and mutual links are suitable to extract semantic relationships of thesaurus. It is needed to apply to update various thesauri, including multilingual thesaurus, in order to generalize the results of this research.

• Journal of the Korean Institute of Oriental Medical Informatics
• /
• v.15 no.2
• /
• pp.1-6
• /
• 2009

The discipline of terminology is based on its own theoretical principles and consists primarily of the following aspects: analysing the concepts and concept structures used in a field or domain of activity, identifying the terms assigned to the concepts, in the case of bilingual or multilingual terminology, establishing correspondences between terms in the various languages, creating new terms, as required. The word properties has syntax, morphology and orthography. The syntax is that how words are put together. The morphology is consist of inflection, derivation, and compounding. The orthography is spelling. Otherwise, the terms of TKM(Traditional Korean Medicine) is two important element of visual character and phonetic notation. A visual character consist of spell, sort words, stop words, etc. For example, that is a case of sort words in which this '다한', '한다', '多汗', '汗多' as same. A phonetic notation consist of palatalization, initial law, etc. For example, that is a case of palatalization in which this '수족랭', '수족냉', '手足冷', '手足冷' as same. Therefore, to enlarged reference terminology is a method by term's normalization. For such a reason, TKM's terms of normalization is necessary.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.167-171
• /
• 2018

개체명 인식은 문서에서 인명, 지명, 기관명 등의 고유한 의미를 나타내는 단위인 개체명을 추출하고, 추출된 개체명의 범주를 결정하는 작업이다. 최근 개체명 인식과 관련된 연구는 입력 데이터의 앞, 뒤를 고려하기 위한 Bi-RNNs와 출력 데이터 간의 전이 확률을 이용한 CRFs를 결합한 방식을 기반으로 다양한 변형의 심층학습 방법론이 제안되고 있다. 그러나 대부분의 연구는 입력 단위를 단어나 형태소로 사용하고 있으며, 성능 향상을 위해 띄어쓰기 정보, 개체명 사전 자질, 품사 분포 정보 등 다양한 정보를 필요로 한다는 어려움이 있다. 본 논문은 기본적인 학습 말뭉치에서 얻을 수 있는 문자 기반의 입력 정보와 Multi-Head Attention을 추가한 Bi-GRU/CRFs을 이용한 다국어 개체명 인식 방법을 제안한다. 한국어, 일본어, 중국어, 영어에 제안 모델을 적용한 결과 한국어와 일본어에서는 우수한 성능(한국어 $F_1$ 84.84%, 일본어 $F_1$ 89.56%)을 보였다. 영어에서는 $F_1$ 80.83%의 성능을 보였으며, 중국어는 $F_1$ 21.05%로 가장 낮은 성능을 보였다.