• Proceedings of the IEEK Conference
• /
• summer
• /
• pp.382-386
• /
• 2004

Recently, with the machine learning trend, most of the machine translation systems on over the world use two syntax tree sets of two relevant languages to learn syntactic tree transfer rules. However, for the English-Vietnamese language pair, this approach is impossible because until now we have not had a Vietnamese syntactic tree set which is correspondent to English one. Building of a very large correspondent Vietnamese syntactic tree set (thousands of trees) requires so much work and take the investment of specialists in linguistics. To take advantage from our available English-Vietnamese Corpus (EVC) which was tagged in word alignment, we choose the SITG (Stochastic Inversion Transduction Grammar) model to construct English- Vietnamese syntactic tree sets automatically. This model is used to parse two languages at the same time and then carry out the syntactic tree transfer. This English-Vietnamese bilingual syntactic tree set is the basic training data to carry out transferring automatically from English syntactic trees to Vietnamese ones by machine learning models. We tested the syntax analysis by comparing over 10,000 sentences in the amount of 500,000 sentences of our English-Vietnamese bilingual corpus and first stage got encouraging result $(analyzed\;about\;80\%)[5].$ We have made use the TBL algorithm (Transformation Based Learning) to carry out automatic transformations from English syntactic trees to Vietnamese ones based on that parallel syntactic tree transfer set[6].

• Journal of the Korea Convergence Society
• /
• v.11 no.11
• /
• pp.33-39
• /
• 2020

In the field of Natural Language Processing, Various research such as Translation, POS Tagging, Q&A, and Sentiment Analysis are globally being carried out. Sentiment Analysis shows high classification performance for English single-domain datasets by pretrained sentence embedding models. In this thesis, the classification performance is compared by Korean E-commerce online dataset with various domain attributes and 6 Neural-Net models are built as BOW (Bag Of Word), LSTM[1], Attention, CNN[2], ELMo[3], and BERT(KoBERT)[4]. It has been confirmed that the performance of pretrained sentence embedding models are higher than word embedding models. In addition, practical Neural-Net model composition is proposed after comparing classification performance on dataset with 17 categories. Furthermore, the way of compressing sentence embedding model is mentioned as future work, considering inference time against model capacity on real-time service.

• Journal of KIISE:Computing Practices and Letters
• /
• v.11 no.1
• /
• pp.92-99
• /
• 2005

In this paper, we present a high-level design methodology applied on a network system-on-a-chip(SOC) using SystemC. The main target of our approach is to get optimum performance parameters for high network address translation(NAT) throughput. The Fast Ethernet media access controller(MAC) and its direct memory access(DMA) controller are modeled with SystemC in transaction level. They are calibrated through the cycle-based measurement of the operation of the real Verilog register transfer language(RTL). The NAT throughput of the model is within $\pm$10% error compared to the output of the real evaluation board. Simulation speed of the model is more than 100 times laster than the RTL. The validated models are used for intensive architecture exploration to find the performance bottleneck in the NAT router.

• Annual Conference on Human and Language Technology
• /
• 1999.10e
• /
• pp.203-208
• /
• 1999

한국어는 잉여성과 중의성의 범 언어적인 특징과 함께 다른 언어에 비해 주어의 생략이 두드러지며 어순이 자유롭기 때문에 구문 형식의 지배를 덜 받는다는 개별적인 특성을 지닌다. 이러한 특성으로 인해 기계번역의 패턴을 추출할 때 서로 유사 가능성이 있는 패턴에 대한 고려가 없이는 같은 의미의 서로 다른 여러 개의 패턴을 모두 하나의 패턴으로 처리하는 오류를 범할 위험이 있다. 본 연구에서 사용되는 구 단위 패턴은 동사구, 명사구, 형용사구 그리고 부사구를 중심으로 한국어 패턴, 패턴 대표 카테고리, 한국어 패턴의 중심어 및 제약조건 대역영어패턴 의미코드로 나뉜다. 범 언어적인 특성의 한국어와 영어간 격차를 해소하기 위해 각각의 명사에 의미코드를 사용하여 다중 언어기반 체계를 구축하였으며. 한국어의 개별적인 특성으로 인해 발생하는 문제를 해소하기 위해 중심어 부과 자질을 사용하였다. 중심어 부과 자질에 있어서, 특히 술어기능명사를 중심어로 하는 기능동사 '하-' 구문은 다른 동사 구문의 형식과는 달리 논항의 수와 형태를 동사가 아닌 명사가 수행하게 된다. 이러한 특징에 대한 변별적인 자질 부여는 구문의 형태-통사적 특징 뿐만이 아니라 의미적인 고유의 특성까지도 잘 뒷받침하면서 패턴 추출에 월등한 효율성을 제시할 수 있다. 향후 이에 대한 연구는 전반적인 기능동사 구문뿐만이 아니라 개별적인 특징을 보이는 모든 구문에 대한 연구로 확대되어 패턴 기반 기계번역의 패턴 추출에 기본적인 정보의 역할을 담당해야 할 것이다.

• 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 Intelligence and Information Systems
• /
• v.10 no.2
• /
• pp.1-13
• /
• 2004

This paper presents a method for the multilingual product retrieval agent through XML and the semantic networks in e-commerce. Retrieval for products is an important process, since it represents interfaces of the customer contact to the e-commerce. Keyword-based retrieval is efficient as long as the product information is structured and organized. But when the product information is expressed across many online shopping malls, especially when it is expressed in different languages with cultural backgrounds, buyers' product retrieval needs language translation with ambiguities resolved in a specific context. This paper presents a RDF modeling case that resolves semantic problems in the representation of product information and across the boundaries of language domains. With adoption of UNSPSC code system, this paper designs and implements an architecture for the multilingual product retrieval agents. The architecture is based on the central repository model of product catalog management with distributed updating processes. It also includes the perspectives of buyers and suppliers. And the consistency and version management of product information are controlled by UNSPSC code system. The multilingual product names are resolved by semantic networks, thesaurus and ontology dictionary for product names.

• The Journal of Bigdata
• /
• v.6 no.1
• /
• pp.63-81
• /
• 2021

The performance of natural language processing is rapidly improving due to the recent development and application of machine learning and deep learning technologies, and as a result, the field of application is expanding. In particular, as the demand for analysis on unstructured text data increases, interest in NLP(Natural Language Processing) is also increasing. However, due to the complexity and difficulty of the natural language preprocessing process and machine learning and deep learning theories, there are still high barriers to the use of natural language processing. In this paper, for an overall understanding of NLP, by examining the main fields of NLP that are currently being actively researched and the current state of major technologies centered on machine learning and deep learning, We want to provide a foundation to understand and utilize NLP more easily. Therefore, we investigated the change of NLP in AI(artificial intelligence) through the changes of the taxonomy of AI technology. The main areas of NLP which consists of language model, text classification, text generation, document summarization, question answering and machine translation were explained with state of the art deep learning models. In addition, major deep learning models utilized in NLP were explained, and data sets and evaluation measures for performance evaluation were summarized. We hope researchers who want to utilize NLP for various purposes in their field be able to understand the overall technical status and the main technologies of NLP through this paper.

• The KIPS Transactions:PartB
• /
• v.11B no.2
• /
• pp.233-240
• /
• 2004

This paper presents a method to implicitly resolve ambiguities using dynamic incremental clustering in Korean-to-English and Japanese-to-English cross-language information retrieval (CLIR). The main objective of this paper shows that document clusters can effectively resolve the ambiguities tremendously increased in translated queries as well as take into account the context of all the terms in a document. In the framework we propose, a query in Korean/Japanese is first translated into English by looking up bilingual dictionaries, then documents are retrieved for the translated query terms based on the vector space retrieval model or the probabilistic retrieval model. For the top-ranked retrieved documents, query-oriented document clusters are incrementally created and the weight of each retrieved document is re-calculated by using the clusters. In the experiment based on TREC test collection, our method achieved 39.41% and 36.79% improvement for translated queries without ambiguity resolution in Korean-to-English CLIR, and 17.89% and 30.46% improvements in Japanese-to-English CLIR, on the vector space retrieval and on the probabilistic retrieval, respectively. Our method achieved 12.30% improvements for all translation queries, compared with blind feedback in Korean-to-English CLIR. These results indicate that cluster analysis help to resolve ambiguity.

• Journal of the Korea Society for Simulation
• /
• v.21 no.1
• /
• pp.55-68
• /
• 2012

Recently it is important to understand service supply chain because the economy moves from manufacturing to services. However, most of existing supply chain research focuses exclusively on the manufacturing sector. To overcome this situation, it needs to investigate and analyze service supply chain. Simulation is one of the most frequently used techniques for analysis and design of complex system. Service supply chain is complex and large systems that require an accurate designing phase. Especially, it is important to examine closely the dynamically interactive behavior of the different service supply chain components in order to predict the performance of the servcie supply chain. In this paper, we develop a conceptual model of service supply chain. Then, we present a new procedure to develop simulation model for the developed conceptual model of service supply chain, based on the UML analysis and design tools and on the ARENA simulation language. The two main characteristics of the proposed procedure are the definition of a systematic procedure to design service supply chain and of a set of rules for the conceptual model translation in an ARENA simulation language. The goal is to improve the knowledge on service supply chain management and support the simulation model development efficiency on service supply chain.

• Journal of the Korea Convergence Society
• /
• v.13 no.3
• /
• pp.11-22
• /
• 2022

Recently, research and development using various Artificial Intelligence (AI) technologies are being conducted in the field of education. Among the AI in Education (AIEd), conversational agents are not limited by time and space, and can learn more effectively by combining them with various AI technologies such as voice recognition and translation. This paper conducted a trend analysis on platforms that have a large number of users and used conversational agents for English learning among commercialized application. Currently commercialized educational platforms using conversational agent through trend analysis has several limitations and problems. To analyze specific problems and limitations, a comparative experiment was conducted with the latest pre-trained large-capacity dialogue model. Sensibleness and Specificity Average (SSA) human evaluation was conducted to evaluate conversational human-likeness. Based on the experiment, this paper propose the need for trained with large-capacity parameters dialogue models, educational data, and information retrieval functions for effective English conversation learning.