• 한국어정보학회:학술대회논문집
• /
• 2016.10a
• /
• pp.176-180
• /
• 2016

대화형 시스템이 사람의 경청 기술을 모방할 수 있다면 대화 상대방과 더 효과적으로 상호작용 할 수 있을 것이다. 본 논문에서는 시스템이 경청 기술을 모방할 수 있도록 사용자의 발화를 기반으로 질문을 생성하는 것에 대해 연구하였다. 그리고 이러한 연구를 위해 필요한 데이터를 Image captioning과 Visual QA 데이터를 기반으로 생성하고 활용하는 방안에 대해 제안한다. 또한 이러한 데이터를 Attention 메커니즘을 적용한 Sequence to sequence 모델에 적용하여 질문을 생성하고, 생성된 질문의 질문 유형을 분석하였다. 마지막으로 사람이 작성한 질문과 모델의 질문 생성 결과 비교를 BLEU 점수를 이용하여 수행하였다.

• Journal of Industrial Technology
• /
• v.21 no.A
• /
• pp.147-153
• /
• 2001

This paper presents techniques for discovering rules with elastic patterns. Elastic patterns are useful for discovering rules from data sequences with different sampling rates. For fast discovery of rules whose heads and bodies are elastic patterns, we construct a suffix tree from succinct forms of data sequences. The suffix tree is a compact representation of rules, and is also used as an index structure for finding rules matched to a target head sequence. When matched rules cannot be found, the concept of rule relaxation is introduced. Using a cluster hierarchy and a relaxation error, we find the least relaxed rules that provide the most specific information on a target head sequence. Performance evaluation through extensive experiments reseals the effectiveness of the proposed approach.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.5 s.43
• /
• pp.321-327
• /
• 2006

Recently there has been enormous growth in the amount of commercial and scientific data. Such datasets consist of sequence data that have an inherent sequential nature. However, only a few of the existing clustering algorithms consider sequentiality. This study presents a similarity measure and a method for clustering such sequence datasets. Especially, we present an extended concept of the measure of similarity, which considers various conditions. Using a splice dataset, we show that the quality of clusters generated by our proposed clustering algorithm is better than that of clusters produced by traditional clustering algorithms.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.176-180
• /
• 2016

대화형 시스템이 사람의 경청 기술을 모방할 수 있다면 대화 상대방과 더 효과적으로 상호작용 할 수 있을 것이다. 본 논문에서는 시스템이 경청 기술을 모방할 수 있도록 사용자의 발화를 기반으로 질문을 생성하는 것에 대해 연구하였다. 그리고 이러한 연구를 위해 필요한 데이터를 Image captioning과 Visual QA 데이터를 기반으로 생성하고 활용하는 방안에 대해 제안한다. 또한 이러한 데이터를 Attention 메커니즘을 적용한 Sequence to sequence 모델에 적용하여 질문을 생성하고, 생성된 질문의 질문 유형을 분석하였다. 마지막으로 사람이 작성한 질문과 모델의 질문 생성 결과 비교를 BLEU 점수를 이용하여 수행하였다.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.32 no.4
• /
• pp.192-202
• /
• 2009

Disassembly of products at their end-of-life (EOL) is a prerequisite for recycling or remanufacturing, since most products should be disassembled before being recycled or remanufactured as secondary parts or materials. In disassembly sequence planning of EOL products, considered are the uncertainty issues, i.e., defective parts or joints in an incoming product, disassembly damage, and imprecise net profits and costs. The paper deals with the problem of determining the disassembly level and corresponding sequence, with the objective of maximizing the overall profit under uncertainties in disassembly cost and/or revenue. The solution is represented as the longest path on a directed acyclic graph where parameter (arc length) uncertainties are modeled in the form of intervals. And, a heuristic algorithm is developed to find a path with the minimum worst case regret, since the problem is NP-hard. Computational experiments are carried out to show the performance of the proposed algorithm compared with the mixed integer programming model and Conde's heuristic algorithm.

• Proceedings of the KIEE Conference
• /
• 1999.07g
• /
• pp.3067-3069
• /
• 1999

In this paper, A bit processor for controlling sequence logic was implemented, using a FPGA. This processor consists of program memory interface. I/O interface, parts for instruction fetch and decode, registers, ALU, program counter and etc. This FPGA is able to execute sequence instruction during program fetch cycle, because of divided bus system, program bus and data bus. Also this bit processor has instructions set that 16bit or 32bit fixed width, so instruction decoding time and data memory interface time was reduced. This FPGA was synthesized by pASIC 2 SpDE and Synplify-Lite synthesis tool of Quick Logic company. The final simulation for worst cases was successfully performed under a Verilog HDL simulation environment. And the FPGA programmed for an 84 pin PLCC package. Finally, the benchmark was performed to prove that Our FPGA has better performance than DSP(TMS320C32-40MHz) for the sequence logic control of PLC.

• The Korean Journal of Applied Statistics
• /
• v.14 no.1
• /
• pp.103-110
• /
• 2001

This paper considers a multistage experiment. Response scales can be same or different from stage to stage. When variables are of nested structure, the response variable at each stage can be defined conditionally. For analysing such data with compound scales, this paper suggests a sequnce of dependence models and shows how to set up a sequence of models for the driver's liscense test data.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.21 no.2
• /
• pp.209-216
• /
• 1997

A feature is well known as a medium to integrate CAD, CAPP and CAM systems. For a part drawing including both simple geometry and compound geometry, a process plan such as the selection of process, machine tool, cutting tool etc. normally needs simple geometry data and non-geometry data of the feature as the input. However, a extended process plan such as the generation of process sequence, operation sequence, jig & fixture, NC program etc. necessarily needs the compound geometry data as well as the simple geometry data and non-geometry data. In this paper, we propose a feature data model according to the result of analyzing necessary data, including the compound geometry data, the simple geometry data and the non-geometry data. Also, an example of the feature data model in milling process planning is described.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.23 no.12
• /
• pp.224-232
• /
• 2009

Power plant start-up is a complicate process involving hundreds of operations that should be performed either automatically or manually. Several major operations should be proceeded in parallel and each major operation is again broken down into detailed operations that must be carried out in a strict sequence. Even though most of the operations are automated, still substantial portions of the operations are carried out manually and the operational status should be monitored by the crew members, which are quite stressful tasks to be performed in real time. In this paper, a data structure called an Event Sequence Monitoring Graph(ESMG) is proposed for monitoring a sequence of events involved in the power plant start-up process. The ESMG is currently being applied to a thermal power plant with a rated output of 500MW. An application example is shown with the boiler feed water pump system start-up process, which exhibits a good potential for future applications.

• The Journal of Society for e-Business Studies
• /
• v.26 no.1
• /
• pp.127-140
• /
• 2021

Recently, sequence data containing time information, such as sensor measurement data and purchase history, has been generated in various applications. So far, many methods for finding sequences that are significantly different from other sequences among given sequences have been proposed. However, most of them have a limitation that they consider only the order of elements in the sequences. Therefore, in this paper, we propose a new anomalous sequence detection method that considers both the order of elements and the time interval between elements. The proposed method uses an extended LSTM autoencoder model, which has an additional layer that converts a sequence into a form that can help effectively learn both the order of elements and the time interval between elements. The proposed method learns the features of the given sequences with the extended LSTM autoencoder model, and then detects sequences that the model does not reconstruct well as anomalous sequences. Using experiments on synthetic data that contains both normal and anomalous sequences, we show that the proposed method achieves an accuracy close to 100% compared to the method that uses only the traditional LSTM autoencoder.