• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.623-626
• /
• 2024

최근 거대 언어 모델의 발전으로 프로그램 합성 분야에서 활용되고 있는 코드 생성 언어 모델의 보안적 측면에 대한 중요성이 부각되고 있다. 그러나, 이를 위해 모델 전체를 재학습하기에는 많은 자원과 시간이 소모된다. 따라서, 본 연구에서는 효율적인 미세조정 방식 중 하나인 프롬프트 튜닝으로 코드 생성 언어 모델이 안전한 코드를 생성할 확률을 높이는 방법을 탐구한다. 또한 이에 따른 기능적 정확성 간의 상충 관계를 분석한다. 실험 결과를 통해 프롬프트 튜닝이 기존 방법에 비해 추가 파라미터를 크게 줄이면서도 보안률을 향상시킬 수 있음을 알 수 있었다. 미래 연구 방향으로는 새로운 조정 손실함수와 하이퍼파라미터 값을 조정하여 성능을 더욱 향상시킬 수 있는지 조사할 것이다. 이러한 연구는 보다 안전하고 신뢰할 수 있는 코드 생성을 위한 중요한 발전을 이끌 수 있을 것으로 기대된다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.4 no.4
• /
• pp.30-39
• /
• 2003

In this paper, we carried out the performance evaluation of HM-Net(Hidden Markov Network) speech recognition system for Korean speech databases. We adopted to construct acoustic models using the HM-Nets modified by HMMs(Hidden Markov Models), which are widely used as the statistical modeling methods. HM-Nets are carried out the state splitting for contextual and temporal domain by PDT-SSS(Phonetic Decision Tree-based Successive State Splitting) algorithm, which is modified the original SSS algorithm. Especially it adopted the phonetic decision tree to effectively express the context information not appear in training speech data on contextual domain state splitting. In case of temporal domain state splitting, to effectively represent information of each phoneme maintenance in the state splitting is carried out, and then the optimal model network of triphone types are constructed by in the parameter. Speech recognition was performed using the one-pass Viterbi beam search algorithm with phone-pair/word-pair grammar for phoneme/word recognition, respectively and using the multi-pass search algorithm with n-gram language models for sentence recognition. The tree-structured lexicon was used in order to decrease the number of nodes by sharing the same prefixes among words. In this paper, the performance evaluation of HM-Net speech recognition system is carried out for various recognition conditions. Through the experiments, we verified that it has very superior recognition performance compared with the previous introduced recognition system.

• Journal of the Korean Society for Library and Information Science
• /
• v.51 no.4
• /
• pp.99-120
• /
• 2017

In order to develop a semiautomatic support system that allows researchers concerned to efficiently analyze the technical trends for the ever-growing industry and market. This paper introduces a couple of Korean sentence generation models that can automatically generate definitional statements as well as descriptions of technical terms and concepts. The proposed models are based on a deep learning model called LSTM (Long Sort-Term Memory) capable of effectively labeling textual sequences by taking into account the contextual relations of each item in the sequences. Our models take technical terms as inputs and can generate a broad range of heterogeneous textual descriptions that explain the concept of the terms. In the experiments using large-scale training collections, we confirmed that more accurate and reasonable sentences can be generated by CHAR-CNN-LSTM model that is a word-based LSTM exploiting character embeddings based on convolutional neural networks (CNN). The results of this study can be a force for developing an extension model that can generate a set of sentences covering the same subjects, and furthermore, we can implement an artificial intelligence model that automatically creates technical literature.

• Journal of Korean Medicine for Obesity Research
• /
• v.17 no.1
• /
• pp.37-45
• /
• 2017

Recently the global epidemic problem of obesity has stimulated intense interest in the study of physiological mechanisms using animal models as a way to gain crucial data required for translation to human studies. Panax ginseng has been reported to have anti-obesity or antidiabetic effects in many animal studies; however, there have been few studies investigating human obesity. Herein, we will assess and examine the evidence supporting the anti-obesity effect of Panax ginseng in animal models with respect to anthropometric and metabolic outcomes. We will include controlled, comparative studies assessing the effect of Panax ginseng in preclinical studies of obesity. Panax ginseng will be administered during or following the induction of experimental obesity. The primary outcome measure will be anthropometric assessment and the secondary outcome measures will include adipose tissue weight, total amount of food consumed and metabolic parameters. We will search MEDLINE, Embase, PubMed, Web of Science, and Scopus without language, publication date, or other restrictions. Ethical approval will not be necessary as the data collected in this study will not be individual patient data, consequently there will be no concerns about violations of privacy. After finishing the whole procedure, the results will be disseminated by publication in a peer-reviewed journal or presented at a relevant conference. This protocol has been registered on the Collaborative Approach to Meta-Analysis and Review of Animal Data from Experimental Studies (CAMARADES) website (http://www.camarades.info).

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.11
• /
• pp.5827-5833
• /
• 2013

Due to science and technology evolutions, modern systems are becoming larger and more complex. In developing complex systems, Model-Based Systems Engineering (MBSE), which is approach to reduce complexity, is being introduced and applied to various system domains. However, because of the modeling being made through a variety of languages, there is a problem with communication within the stakeholders and a lack of consistency in the models. In this paper, by investigating the rule explaining the transformation of one of the only traditional diagrams, DFD, to SysML and reusing the formerly built models, we attempt to implement by SysML. Analyzing each diagram's Metamodel and validating the connection of each component through bipartite graph especially suggest an effective transformation rule. Also, by applying to naval-combat system, we confirm efficiency of this study. Establishing the results of this study as basis for conducting further study, we will be able to transform other previous models gained from formerly built system to SysML. In this way, the stakeholder's communication can be improved and we anticipate that the application of SysML will be beneficial to the much efficient MBSE.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.9
• /
• pp.308-313
• /
• 2016

Modern systems have become more and more complex due to the ever-increasing user requirements and rapid advance of technology. As such, the frequency of accidents due to system design errors or failure has been increasing. When the damage incurred by accidents to human beings or property is serious, the underlying systems are referred to as safety-critical systems. The development of such systems requires special efforts to ensure the safety of the human beings operating them. To cope with such a requirement, in this paper an approach is employed in which we consider safety starting from the conceptual design phase of the systems. Specifically, a systems design method that can detect functional failure is proposed by utilizing meta-models and M&S methods. To accomplish this, the safety design data from international safety standards are first extracted and also a meta-model is generated using SysML (systems modeling language). Then, a SysML-based system design method is proposed based on the use of the developed meta-model. We also discuss how the safety requirements can be created and verified using a simulation method. Finally, through a case study in automotive design, it is demonstrated that the detection of a functional failure and the verification of a safety requirement can be accomplished using the SysML-based M&S method. This study indicates that the use of meta-models can be useful for collecting and managing safety data and that the meta-model based M&S method can make it possible to satisfy the system requirements by reducing the design errors.

• Korean Journal of Logic
• /
• v.22 no.1
• /
• pp.43-86
• /
• 2019

If 1st order ZFC is consistent(has a model($M_1$)) it has a transitive denumerable model($M_2$). This leads to a paradoxical situation called 'Skolem paradox'. This can be easily resolved by Skolem's typical resolution. but In the process, we must accept the model theoretic relativity for the concept of set. This relativity can generate a situation where the meaning of the set concept, for example, is given differently depending on the two models. The problem is next. because the sentence '¬denu(PN)' which indicate that PN is not denumerable is equally true in two models, A indistinguishability problem that the concept <¬denu> is not formally indistinguishable in ZFC arise. First, I will give a detail analysis of what the nature of this problem is. And I will provide three ways of responding to this problem from the standpoint of supporting ZFC. First, I will argue that <¬denu> concept, which can be relative to the different models, can be 'almost' distinguished in ZFC by using the formalization of model theory in ZFC. Second, I will show that <¬denu> can change its meaning intrinsically or naturally, by its contextual dependency from the semantic considerations about quantifier that plays a key role in the relativity of <¬denu>. Thus, I will show the model-relative meaning change of <¬denu> concept is a natural phenomenon external to the language, not a matter of responsible for ZFC.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.5
• /
• pp.530-536
• /
• 2021

Recently, due to the development of deep learning, end-to-end speech recognition, which directly maps graphemes to speech signals, shows good performance. Especially, among the end-to-end models, conformer shows the best performance. However end-to-end models only focuses on the probability of which grapheme will appear at the time. The decoding process uses a greedy search or beam search. This decoding method is easily affected by the final probability output by the model. In addition, the end-to-end models cannot use external pronunciation and language information due to structual problem. Therefore, in this paper conformer with lexicon transducer is proposed. We compare phoneme-based model with lexicon transducer and grapheme-based model with beam search. Test set is consist of words that do not appear in training data. The grapheme-based conformer with beam search shows 3.8 % of CER. The phoneme-based conformer with lexicon transducer shows 3.4 % of CER.

• Journal of Internet Computing and Services
• /
• v.25 no.4
• /
• pp.47-56
• /
• 2024

This paper proposes a method for fine-tuning BERT-based natural language processing models to perform sentiment analysis on Korean review data. By varying the input sequence length during this process and comparing the performance, we aim to explore the optimal performance according to the input sequence length. For this purpose, text review data collected from the clothing shopping platform M was utilized. Through web scraping, review data was collected. During the data preprocessing stage, positive and negative satisfaction scores were recalibrated to improve the accuracy of the analysis. Specifically, the GPT-4 API was used to reset the labels to reflect the actual sentiment of the review texts, and data imbalance issues were addressed by adjusting the data to 6:4 ratio. The reviews on the clothing shopping platform averaged about 12 tokens in length, and to provide the optimal model suitable for this, five BERT-based pre-trained models were used in the modeling stage, focusing on input sequence length and memory usage for performance comparison. The experimental results indicated that an input sequence length of 64 generally exhibited the most appropriate performance and memory usage. In particular, the KcELECTRA model showed optimal performance and memory usage at an input sequence length of 64, achieving higher than 92% accuracy and reliability in sentiment analysis of Korean review data. Furthermore, by utilizing BERTopic, we provide a Korean review sentiment analysis process that classifies new incoming review data by category and extracts sentiment scores for each category using the final constructed model.

• Journal of Intelligence and Information Systems
• /
• v.26 no.3
• /
• pp.109-125
• /
• 2020

The Ministry of National Defense is pushing for the Defense Acquisition Program to build strong defense capabilities, and it spends more than 10 trillion won annually on defense improvement. As the Defense Acquisition Program is directly related to the security of the nation as well as the lives and property of the people, it must be carried out very transparently and efficiently by experts. However, the excessive diversification of laws and regulations related to the Defense Acquisition Program has made it challenging for many working-level officials to carry out the Defense Acquisition Program smoothly. It is even known that many people realize that there are related regulations that they were unaware of until they push ahead with their work. In addition, the statutory statements related to the Defense Acquisition Program have the tendency to cause serious issues even if only a single expression is wrong within the sentence. Despite this, efforts to establish a sentence comparison system to correct this issue in real time have been minimal. Therefore, this paper tries to propose a "Comparison System between the Statement of Military Reports and Related Laws" implementation plan that uses the Siamese Network-based artificial neural network, a model in the field of natural language processing (NLP), to observe the similarity between sentences that are likely to appear in the Defense Acquisition Program related documents and those from related statutory provisions to determine and classify the risk of illegality and to make users aware of the consequences. Various artificial neural network models (Bi-LSTM, Self-Attention, D_Bi-LSTM) were studied using 3,442 pairs of "Original Sentence"(described in actual statutes) and "Edited Sentence"(edited sentences derived from "Original Sentence"). Among many Defense Acquisition Program related statutes, DEFENSE ACQUISITION PROGRAM ACT, ENFORCEMENT RULE OF THE DEFENSE ACQUISITION PROGRAM ACT, and ENFORCEMENT DECREE OF THE DEFENSE ACQUISITION PROGRAM ACT were selected. Furthermore, "Original Sentence" has the 83 provisions that actually appear in the Act. "Original Sentence" has the main 83 clauses most accessible to working-level officials in their work. "Edited Sentence" is comprised of 30 to 50 similar sentences that are likely to appear modified in the county report for each clause("Original Sentence"). During the creation of the edited sentences, the original sentences were modified using 12 certain rules, and these sentences were produced in proportion to the number of such rules, as it was the case for the original sentences. After conducting 1 : 1 sentence similarity performance evaluation experiments, it was possible to classify each "Edited Sentence" as legal or illegal with considerable accuracy. In addition, the "Edited Sentence" dataset used to train the neural network models contains a variety of actual statutory statements("Original Sentence"), which are characterized by the 12 rules. On the other hand, the models are not able to effectively classify other sentences, which appear in actual military reports, when only the "Original Sentence" and "Edited Sentence" dataset have been fed to them. The dataset is not ample enough for the model to recognize other incoming new sentences. Hence, the performance of the model was reassessed by writing an additional 120 new sentences that have better resemblance to those in the actual military report and still have association with the original sentences. Thereafter, we were able to check that the models' performances surpassed a certain level even when they were trained merely with "Original Sentence" and "Edited Sentence" data. If sufficient model learning is achieved through the improvement and expansion of the full set of learning data with the addition of the actual report appearance sentences, the models will be able to better classify other sentences coming from military reports as legal or illegal. Based on the experimental results, this study confirms the possibility and value of building "Real-Time Automated Comparison System Between Military Documents and Related Laws". The research conducted in this experiment can verify which specific clause, of several that appear in related law clause is most similar to the sentence that appears in the Defense Acquisition Program-related military reports. This helps determine whether the contents in the military report sentences are at the risk of illegality when they are compared with those in the law clauses.