• Journal of Intelligence and Information Systems
• /
• v.28 no.3
• /
• pp.45-55
• /
• 2022

Recently, large-scale language models (LPLM) have been shown state-of-the-art performances in various tasks of natural language processing including intent classification. However, fine-tuning LPLM requires much computational cost for training and inference which is not appropriate for dialog system. In this paper, we propose compressed intent classification model for multi-agent in low-resource like CPU. Our method consists of two stages. First, we trained sentence encoder from LPLM then compressed it through knowledge distillation. Second, we trained agent-specific adapter for intent classification. The results of three intent classification datasets show that our method achieved 98% of the accuracy of LPLM with only 21% size of it.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.19 no.40
• /
• pp.143-155
• /
• 1996

Presented in this paper are a CAPP(Computer Automated Process Planning) scheme and a generating method of intelligent NC data for unmaned machining of mold die. Mold die surfaces usually have free-formed geometry of complex shapes. So it is easy to overcut the die surface and to overload the cutting tools. It takes tens of hours to prepare process plans and to generate NC data for each processes. Therefore a classification of unit machining operation(UMO) for mold die manufacture, a backward recursive capp algorithm and a generating method of intelligent NC data are presented in this paper in order to provide a unmaned machining architecture of mold die.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2001.12a
• /
• pp.291-294
• /
• 2001

최근 들어 개인용 컴퓨터의 광범위한 보급과 인터넷의 확산으로 인하여 이러한 시스템을 사용하는 사용자들은 보다 편리하고 사용자의 요구에 적절하게 대응할 수 있는 지능화 된 시스템을 필요로 하게 되었다. 이러한 배경을 바탕으로 하여 지능형 에이전트 이론을 활용한 연구와 개발은 다방면에 걸쳐 활성화되고 있다. 본 연구에서는 패턴분류에 있어서의 몇 가지 알고리즘을 통하여 공통된 데이터에 대한 패턴 클러스터링을 통한 패턴의 분류 방법을 고찰하고, 또한 지능형 에이전트 개념을 적용하여 패턴분류를 위한 지능형 에이전트 시스템을 모델링하고 구현하였다. 그 결과 4개의 알고리즘에 대한 300개의 3차원 데이터의 패턴분류가 정확하게 되는 것을 확인하였으며, 본 연구의 핵심 분야인 지능형 에이전트 시스템의 다양한 에이전트들을 적용하여 기존의 시스템과는 차별화 된 인터페이스가 이루어질 수 있음을 보인다.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.653-657
• /
• 2005

AGV is very useful equipment to transfer containers in automated container terminal. AGV must have Obstacle Detection System (ODS) for port automation. ODS needs the function to classify some specified object from background in acquired data. And it must be able to track classified moving objects. Finally, ODS could determine its next action for safe driving whether it should do emergency stop or speed down, or it should change its deriving lane. For these functions, ODS can have many different kinds of algorithm. In this paper, we present one of AGV to be used in automated container terminal.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2006.05a
• /
• pp.374-377
• /
• 2006

본 논문은 역설계 기법을 이용하여 비접촉 방법으로 인체의 발을 3차원으로 복원하고, 발에 관련된 분야에서 필요한 데이터를 추출하는 시스템에서 신발 제작에 필요한 데이터를 지능 기법을 이용하여 분류하는 모듈을 제안한다. 신발의 경우 개개인의 신체조건이 다르고 유행과 개성을 추구하고자 하기 때문에 기존의 생산체계로는 한계가 있다. 측정데이터를 기반으로 하는 맞춤 신발은 기존의 전통적인 수제화 방식과 대량생산 방식의 장점만을 취하여 저렴하고 편리하게 제작된다. 또한 3차원 측정기를 이용하여 측정한 화형 데이터를 적당하게 분류한다면 기성화와 수제화 제작에 필요한 라스트 생성과 개인의 발의 구조 분석에 활용 가능 할 것이다. 따라서 본 논문에서는 획득된 발 데이터를 미리 정해 놓은 그룹으로 결정하기위해 신경망을 사용하여 신발 제작에 필요한 최적의 라스트 데이터를 선택하게 한다.

• Proceedings of the KIEE Conference
• /
• 1987.11a
• /
• pp.435-438
• /
• 1987

In this paper, among the robot programming languages that enable processing of sensory information, eight exemplary languages are chosen, and investigated in terms of their characteristics, why they are designed the way they are, and the kind of sensors each language can use and apply to. In addition, the characteristics of each language is compared with one another from the sensor point of view and the flow of each language is analyzed from the robot language classification point of view. Finally, We investigate the progress and the requirements of the sensor-based robot programming languages for further developments.

• Journal of Intelligence and Information Systems
• /
• v.10 no.2
• /
• pp.79-89
• /
• 2004

Since the amount of E-mail messages has increased , we need a new technique for efficient e-mail classification. E-mail classifications are grouped into two classes: binary classification, multi-classification. The current binary classification methods are mostly spm mail classification methods which are based on rule driven, bayesian, SVM, etc. The current multi- classification methods are based on clustering which groups e-mails by similarity. In this paper, we propose a novel method for e-mail classification. It combines the automatic category generation method based on the vector model and the dynamic category hierarchy construction method. This method can multi-classify e-mail automatically and manage a large amount of e-mail efficiently. In addition, this method increases the search accuracy by dynamic reclassification of e-mails.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.10
• /
• pp.5132-5148
• /
• 2017

Cyber attacks are evolving commensurate with recent developments in information security technology. Intrusion detection systems collect various types of data from computers and networks to detect security threats and analyze the attack information. The large amount of data examined make the large number of computations and low detection rates problematic. Feature selection is expected to improve the classification performance and provide faster and more cost-effective results. Despite the various feature selection studies conducted for intrusion detection systems, it is difficult to automate feature selection because it is based on the knowledge of security experts. This paper proposes a feature selection technique to overcome the performance problems of intrusion detection systems. Focusing on feature selection, the first phase of the proposed system aims at constructing a feature subset using a sequential forward floating search (SFFS) to downsize the dimension of the variables. The second phase constructs a classification model with the selected feature subset using a random forest classifier (RFC) and evaluates the classification accuracy. Experiments were conducted with the NSL-KDD dataset using SFFS-RF, and the results indicated that feature selection techniques are a necessary preprocessing step to improve the overall system performance in systems that handle large datasets. They also verified that SFFS-RF could be used for data classification. In conclusion, SFFS-RF could be the key to improving the classification model performance in machine learning.

• Journal of Korea Multimedia Society
• /
• v.6 no.5
• /
• pp.824-830
• /
• 2003

RPA (Recursive Partition Averaging) method was proposed in order to improve the storage requirement and classification rate of the Memory Based Reasoning. That algorithm worked well in many areas, however, the major drawbacks of RPA are it's pattern averaging mechanism. We propose an adaptive OHC algorithm which uses the FPD(Feature-based Population Densimeter) to increase the classification rate of RPA. The proposed algorithm required only approximately 40% of memory space that is needed in k-NN classifier, and showed a superior classification performance to the RPA. Also, by reducing the number of stored patterns, it showed a excellent results in terms of classification when we compare it to the k-NN.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.4
• /
• pp.1045-1050
• /
• 1998

Classification techniques are often an importand component of intelligent systems and are use for both deta preprocessing and decision making. In the design of a classification system, the labled samples must be given to provide a priori information for the classification. Moreover, the number of classes to be categorized must be known a priori information, called OFCAM. In OFCAM, an unsupervised by OFCAM, the database of a classification system, called PCSDB, is constructed. Then, PCSDB can be effectively used in the decision process of the system.