• Proceedings of the KIEE Conference
• /
• 2003.11c
• /
• pp.836-839
• /
• 2003

In this paper, we implement a language model by a bigram and evaluate proper smoothing technique for unit of low perplexity. Word, morpheme, clause units are widely used as a language processing unit of the language model. We propose VCCV units which have more small vocabulary than morpheme and clauses units. We compare the VCCV units with the clause and the morpheme units using the perplexity. The most common metric for evaluating a language model is the probability that the model assigns the derivative measures of perplexity. Smoothing used to estimate probabilities when there are insufficient data to estimate probabilities accurately. In this paper, we constructed the N-grams of the VCCV units with low perplexity and tested the language model using Katz, Witten-Bell, absolute, modified Kneser-Ney smoothing and so on. In the experiment results, the modified Kneser-Ney smoothing is tested proper smoothing technique for VCCV units.

• ETRI Journal
• /
• v.46 no.1
• /
• pp.137-153
• /
• 2024

This study introduces a new approach in Korean morphological analysis combining dictionary-based techniques with Transformer-based deep learning models. The key innovation is the use of a BERT-based reranking system, significantly enhancing the accuracy of traditional morphological analysis. The method generates multiple suboptimal paths, then employs BERT models for reranking, leveraging their advanced language comprehension. Results show remarkable performance improvements, with the first-stage reranking achieving over 20% improvement in error reduction rate compared with existing models. The second stage, using another BERT variant, further increases this improvement to over 30%. This indicates a significant leap in accuracy, validating the effectiveness of merging dictionary-based analysis with contemporary deep learning. The study suggests future exploration in refined integrations of dictionary and deep learning methods as well as using probabilistic models for enhanced morphological analysis. This hybrid approach sets a new benchmark in the field and offers insights for similar challenges in language processing applications.

• The KIPS Transactions:PartD
• /
• v.9D no.6
• /
• pp.991-998
• /
• 2002

With the advent of XML and database languages armed with the object-oriented concept and deductive logic, the problem of efficient query processing for them has become a major issue. We describe a way of processing semi-structured XML data through an implementation of a Deductive and Object-oriented Database (DOODB) language with the explanation of query processing. We have shown how to convert an XML data model to a DOODB data model. We have then presented an efficient query processing method based on Connection Graph Resolution. We also present a knowledge-based query processing method that uses the homomorphism of objects in the database and the associative rule of substitutions.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.7
• /
• pp.1941-1954
• /
• 1999

This paper presents a design and implementation of CRs(CHILL Run-time support System) to adapt the concurrent processing facilities of CHILL(CCITT High Level Language) which had recommended by ITU-T(International Telecommunication Union Telecommunication Standardization Sector). Because the CHILL provides more various concurrent processing facilities that other concurrent programming language, a design and implementation on CRS can give us real effects to gain the major functionalities and the techniques of the concurrent processing. In this paper, we design the interface rules of concurrent functions to conform with the CHILL compiler. We use the concurrent processing primitives as the library style to be invoked by procedure calls, and implement the start-up routine of the CHILL program, the context switching routine, and the CHILL process control parts to control be execution of the CHILL processes concurrently.

• Proceedings of the KSMRM Conference
• /
• 2006.11a
• /
• pp.83-83
• /
• 2006

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.3
• /
• pp.417-428
• /
• 1996

In this paper, a database interface system with natural language dialogue is designed and implemented. The system is made up of language analysis, context processing, dialogue processing and DB processing unit. The method for classifying and processing an undefined word in language analysis is proposed. It reduces the dictionary size, which gives difficulties in DB Interface. And the current DB Interfaces dealt with an input utterance independently. But the system in this paper provides a user with the interface environment in which he or she can have a continuous conversation with the system and retrieve DB information. Thus in this paper, speech acts which include user's inattentions well as propositional contents are defined, and user action hierarchical model for library DB retrieval is constructed. And the system uses the defined knowledge to recognize-user's plan, effectively understanding and managing the ongoing dialogue. And the system is implemented in the domain of library database in order to prove the proposed methods in this paper.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.4
• /
• pp.115-119
• /
• 2023

The natural language processing models used in current artificial intelligence are huge, causing various difficulties in processing and analyzing data in real time. In order to solve these difficulties, we proposed a method to improve the efficiency of processing by using less memory and checked the performance of the proposed model. The technique applied in this paper to evaluate the performance of the proposed model is to divide the large corpus by adjusting the number of attention heads and embedding size of the BERT[1] model to be small, and the results are calculated by averaging the output values of each forward. In this process, a random offset was assigned to the sentences at every epoch to provide diversity in the input data. The model was then fine-tuned for classification. We found that the split processing model was about 12% less accurate than the unsplit model, but the number of parameters in the model was reduced by 56%.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.174-180
• /
• 2024

In this paper, we explore an emotion classification method through multimodal learning utilizing wav2vec 2.0 and KcELECTRA models. It is known that multimodal learning, which leverages both speech and text data, can significantly enhance emotion classification performance compared to methods that solely rely on speech data. Our study conducts a comparative analysis of BERT and its derivative models, known for their superior performance in the field of natural language processing, to select the optimal model for effective feature extraction from text data for use as the text processing model. The results confirm that the KcELECTRA model exhibits outstanding performance in emotion classification tasks. Furthermore, experiments using datasets made available by AI-Hub demonstrate that the inclusion of text data enables achieving superior performance with less data than when using speech data alone. The experiments show that the use of the KcELECTRA model achieved the highest accuracy of 96.57%. This indicates that multimodal learning can offer meaningful performance improvements in complex natural language processing tasks such as emotion classification.

• Annual Conference on Human and Language Technology
• /
• 1998.10c
• /
• pp.329-334
• /
• 1998

자연언어 처리 시스템은 시제품 개발이 비교적 용이한 반면에 이를 실용적인 시스템으로 발전 시키는데 많은 어려움이 있다. 본 논문에서는 형태소 분석기와 구문분석기, 기계번역 시스템 등을 개발할 때 발생하는 문제점과 방법론을 고찰하고, 실용적인 시스템을 개발하기 위한 효율적인 방법으로 2-step 패러다임과 이를 실현하는 방안으로 기능별 모듈화에 의한 divide and conquer 기법, 단순화 기법, 예외처리 기법 등을 활용하는 방법을 제안한다.

• Annual Conference on Human and Language Technology
• /
• 1996.10a
• /
• pp.346-350
• /
• 1996

한일 양언어의 복합문은 여러가지의 통어 현상을 가지며, 주어, 목적어 등의 생략 현상으로 문장의 표층상에 나타나지 않는 것이 있기 때문에 수식구조의 처리가 복잡해지고, 구문해석에 있어서 애매성의 요인이 된다. 따라서, 본 논문에서는 DPN에 의하여 한국어와 일본어의 수식 scope를 해석하는 방법에 대하여 설명한다. 먼저, 한일 양언어의 공통점과 차이점을 찾아내어, 한국어와 일본어의 보문을 표현형식으로 나타내고, 동사의 격정보로부터 DPN을 구성하여 DPN상에서 보문의 수식 Scope를 해석하는 방법에 대해서 설명한다.