This paper reports recent advances in automatic speech summarization method. In our proposed method, a set of words maximizing a summarization score is extracted from automatically transcribed speech. This extraction ...
详细信息
ISBN:
(纸本)8790834100
This paper reports recent advances in automatic speech summarization method. In our proposed method, a set of words maximizing a summarization score is extracted from automatically transcribed speech. This extraction is performed according to a target compression ratio using a dynamic programming technique. The extracted set of words is then connected to build a summarized sentence. The summarization score consists of a word significance measure, a confidence measure, linguistic likelihood, and a word concatenation probability which is determined by a dependency structure in the original speech given by Stochastic Dependency Context Free Grammar (SDCFG). Japanese broadcast news speech transcribed using a large vocabulary continuous speech recognition (LVCSR) system is summarized using our proposed method and evaluated in comparison with manual summarization by human subjects. The manual summarization results are combined to build a word network, and word accuracy of each automatic summarization result is calculated comparing with the most similar word string in the network.
The problem of databases containing missing values is a common one in the medical environment. Researchers must find a way to incorporate the incomplete data into the data set to use those cases in their experiments. ...
详细信息
Many real world problems deal with ordering objects instead of classifying objects, although majority of research in machine learning and data mining has been focused on the latter. For modeling ordering problems, we ...
详细信息
ISBN:
(纸本)0769511198
Many real world problems deal with ordering objects instead of classifying objects, although majority of research in machine learning and data mining has been focused on the latter. For modeling ordering problems, we generalize the notion of information tables to ordered information tables by adding order relations on attribute values. The problem of mining ordering rules is formulated as finding association between orderings of attribute values and the overall ordering of objects. An ordering rules may state that "if the value of an object x on an attribute a is ordered ahead of the value of another object y on the same attribute, then x is ordered ahead of y". For mining ordering rules, we first transform an ordered information table into a binary information, and then apply any standard machine learning and data mining algorithms. As an illustration, we analyze in detail Maclean's universities ranking for the year 2000.
Rough sets have traditionally been applied to decision (classification) problems. We suggest that rough sets are even better suited for reasoning. It has already been shown that rough sets can be applied for reasoning...
详细信息
In order to detect small reverse-side cracks existing on 5 mm thick steel plates, we have measured strength of residual magnetization at 1 mm above 5 mm thick steel plates. A differential type magneto-impedance effect...
In order to detect small reverse-side cracks existing on 5 mm thick steel plates, we have measured strength of residual magnetization at 1 mm above 5 mm thick steel plates. A differential type magneto-impedance effect sensor (a gradio-MI sensor) is applied to measure it. Because the gradio-MI sensor is suitable for detection of localized micro magnetic field. From results of our experiments, a distribution of the residual differential magnetic field at 1 mm above a specimen is clearly changed near a reverse-side crack of 1 mm depth with 5 mm length. In this paper, experimental results utilizing the distribution of a residual differential magnetic field measured by gradio-MI sensors is shown. And, validity of the residual magnetic field method for inspection of small reverse-side cracks on thick steel plates is discussed.
作者：
BYUNG-JU KANGKEY-SUN CHOIDivision of Computer Science
Department of Electrical Engineering & Computer Science Korea Terminology Research Center for Language and Knowledge Engineering (KORTERM) Advanced Information Technology Research Center (AITrc) Korea Advanced Institute of Science and Technology (KAIST) 373-1 Kusong-dong Yusong-gu Taejon 305-701 Korea
In Korean text these days, the use of English words with or without phonetic translations are growing at a high speed. To make matters worse, the Korean transliteration of an English word may vary greatly. The mixed u...
详细信息
In Korean text these days, the use of English words with or without phonetic translations are growing at a high speed. To make matters worse, the Korean transliteration of an English word may vary greatly. The mixed use of English words and their various transliterations in the same document or document collection may cause severe word mismatch problems in Korean information retrieval. There are two possible approaches to tackle this problem: transliteration and back-transliteration method. We argue that our newly proposed transliteration approach is more advantageous for the resolution of the word mismatch problem than the previously proposed back-transliteration approach. Our information retrieval experiment results support this argument.
A modeling based on integral equation methods is shown to investigate the effect of material anisotropy on an angle beam ultrasonic testing. Scattered amplitudes are numerically obtained to simulate pulse echoes measu...
A modeling based on integral equation methods is shown to investigate the effect of material anisotropy on an angle beam ultrasonic testing. Scattered amplitudes are numerically obtained to simulate pulse echoes measured at various surface points of an anisotropic steel plate with artificial cylindrical cavities. A good agreement between calculation and experiment is obtained. The proposed modeling and analysis will be useful to correct quantitatively the effect of material anisotropy in a ultrasonic testing.
In this paper, we develop a dynamics-based adaptive control scheme to achieve the stability for teleoperation system and its transparency in the sense of motion/force tracking. The proposed scheme does not require acc...
详细信息
In this paper, we develop a dynamics-based adaptive control scheme to achieve the stability for teleoperation system and its transparency in the sense of motion/force tracking. The proposed scheme does not require accurate dynamic parameters of manipulators, human and environment models as well as the information of manipulators' acceleration. The validity and advantages on control characteristics of proposed scheme towards conventional impedance matching control are confirmed by numerical simulations and experiments.
PID controllers have been widely used in many chemical processes. Because they have only three control parameters and their physical meanings can be easily grasped. However, it is difficult to tune those parameters pr...
详细信息
PID controllers have been widely used in many chemical processes. Because they have only three control parameters and their physical meanings can be easily grasped. However, it is difficult to tune those parameters practically, since the process dynamics often change due to operating conditions or various disturbances. For this problem, a design method of robust PID controllers has been already proposed by the authors in order to guarantee the stability of the control system. However, as that control method is conservative, the desirable setpoint response can not be always obtained. In this paper, a design scheme of a self-tuning pre-filter is proposed to supplement the robust PID controller based on the two-degree-of-freedom control scheme. According to the proposed scheme, the transient property for the setpoint response can be improved keeping the robust stability. Finally, the proposed scheme is experimentally evaluated on an air pressure control system.
Much research on designing self-tuning control systems for linear systems have been proposed by using the least squares parameter identification method. However, it is difficult to employ the algorithm for the nonline...
详细信息
Much research on designing self-tuning control systems for linear systems have been proposed by using the least squares parameter identification method. However, it is difficult to employ the algorithm for the nonlinear systems except for the case where the unknown parameters are linearly combined with nonlinear terms. In this paper, a parameter estimation scheme for nonlinear systems is proposed by using a neural network. Furthermore, a design method of the control system is derived by minimizing a cost function of the generalized minimum variance control. This control input is calculated by using the estimated parameters. Finally, the effectiveness of the proposed scheme is numerically evaluated.
暂无评论