• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2022년도 추계학술발표대회
• /
• pp.308-310
• /
• 2022

With the rise of the Al wave, task-oriented dialogue systems have become one of the popular research directions in academia and industry. Currently, task-oriented dialogue systems mainly adopt pipelined form, which mainly includes natural language understanding, dialogue state decision making, dialogue state tracking and natural language generation. However, pipelining is prone to error propagation, so many task-oriented dialogue systems in the market are only for single-round dialogues. Usually single- domain dialogues have relatively accurate semantic understanding, while they tend to perform poorly on multi-domain, multi-round dialogue datasets. To solve these issues, we developed a paddle-based multi-domain task-oriented Chinese dialogue system. It is based on NEZHA-base pre-training model and CrossWOZ dataset, and uses intention recognition module, dichotomous slot recognition module and NER recognition module to do DST and generate replies based on rules. Experiments show that the dialogue system not only makes good use of the context, but also effectively addresses long-term dependencies. In our approach, the DST of dialogue tracking state is improved, and our DST can identify multiple slotted key-value pairs involved in the discourse, which eliminates the need for manual tagging and thus greatly saves manpower.

• 전자공학회논문지B
• /
• 제33B권4호
• /
• pp.131-140
• /
• 1996

In this paper, we propose the method that extracts the speech feature using the hearing model through signal precessing techniques. The proposed method includes following procedure ; normalization of the short-time speech block by its maximum value, multi-resolution analysis using the discrete wavelet transformation and re-synthesize using thediscrete inverse wavelet transformation, differentiation after analysis and synthesis, full wave rectification and integration. In order to verify the performance of the proposed speech feature in the speech recognition task, korean digita recognition experiments were carried out using both the dTW and the VQ-HMM. The results showed that, in case of using dTW, the recognition rates were 99.79% and 90.33% for speaker-dependent and speaker-independent task respectively and, in case of using VQ-HMM, the rate were 96.5% and 81.5% respectively. And it indicates that the proposed speech feature has the potentials to use as a simple and efficient feature for recognition task.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 1996년도 10월 학술대회지
• /
• pp.435-447
• /
• 1996

In this paper, we propose the method that extracts the speech feature using the hearing model through signal processing techniques. The proposed method includes the following procedure ; normalization of the short-time speech block by its maximum value, multi-resolution analysis using the discrete wavelet transformation and re-synthesize using the discrete inverse wavelet transformation, differentiation after analysis and synthesis, full wave rectification and integration. In order to verify the performance of the proposed speech feature in the speech recognition task, korean digit recognition experiments were carried out using both the DTW and the VQ-HMM. The results showed that, in the case of using DTW, the recognition rates were 99.79% and 90.33% for speaker-dependent and speaker-independent task respectively and, in the case of using VQ-HMM, the rate were 96.5% and 81.5% respectively. And it indicates that the proposed speech feature has the potential for use as a simple and efficient feature for recognition task

• 인지과학
• /
• 제23권3호
• /
• pp.389-408
• /
• 2012

감각적 경험을 언어로 기술하면 그 경험에 관한 감각적 기억이 저하되는데, 이러한 현상을 일컬어 언어 장막(verbal overshadowing) 효과라고 한다. Schooler(2002)[1]는 언어화로 인해 정보처리 방식이 전역적 처리에서 국지적 처리로 전환되기 때문에 언어 장막이 발생한다고 제안하였다. 본 연구에서는 이러한 정보처리의 전환이 실제로 일어나는가를 검증하고자 얼굴 자극에 대한 기억을 요구하고 이를 토대로 언어적 묘사를 한 직후에 전역 처리와 국지 처리를 비교할 수 있는 Navon 과제를 실시하였다. 얼굴 재인 과제에서는 참가자들에게 외워야 하는 얼굴을 제시한 후 그 얼굴에 대한 언어화를 요구하였으며, 얼굴 재인 과제를 수행한 직후에 Navon 과제를 수행하도록 하였다. 그 결과, 얼굴 재인 과제에서는 언어화를 한 묘사 집단에서 언어화를 하지 않은 통제 집단보다 낮은 재인률을 보이는 언어 장막 효과가 나타났다. Navon 과제에서는 통제 집단이 전역수준의 정보가 국지 수준의 정보보다 우세하게 처리되는 전역 선행성(global precedence)을 보인 반면에, 묘사 집단은 전역 수준의 정보보다 국지 수준의 정보를 더 우세하게 처리하는 국지 선행성을 보였다. 이는 얼굴을 언어화함으로써 정보처리의 방식이 전역적인 방향에서 국지적인 방향으로 전환되고, 그 결과로서 얼굴에 대한 재인이 손상된 것임을 시사한다.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 1999년도 춘계학술발표논문집 논문집
• /
• pp.127-132
• /
• 1999

Cardiovascular, respiratory and electrodermal responses to acute stress episodes modeled by combined presentation of intense white noise and performance of word recognition task with noise background were studied in 15 college students. Experimental procedure consisted in sessions with white noise, word recognition task presentation with noise background and test with noise background. Recorded physiological variables were analyzed in terms of their sensitivity to detect activation of sympathetic and parasympathetic branches of autonomic nervous system and thus reflect autonomic arousal level during shout-term stress-inducing experimental manipulations. It was shown that performance of effortful mental task with noise background elicited significant physiological responses typical for active coping behavior, namely electrodermal arousal and increased cardiovascular activity. this response profile was more profound as compared to white noise only or attending task in noise background. However, all physiological responses were mostly phasic, without long-term tonic changes, since almost all variables recovered to their initial baseline levels, suggesting that dominant autonomic mechanisms in transient acute stress episodes were of parasympathetic nature (withdrawal in stress with subsequent activation in restoration period), while sympathetic contribution was not long-lasting. Nevertheless, increased number of stressors and their longer exposure may result in higher profile of tonic sympathetic arousal and reduced functional role of vagal mechanisms in autonomic balance regulation.

• 대한간호학회지
• /
• 제30권5호
• /
• pp.1170-1180
• /
• 2000

This study was a quasi-experimental study to confirm the effects of a memory training program using efficacy sources. The purpose was to develop an effective memory training program for elderly people and to identify the effects of the memory training program. This study was carried out between February 24 and July 18, 1999 and the subjects of the study were 102 elderly people who were participants at a welfare institute in Seoul. The experimental group (51) and the control group (51) were assigned by means of participation order. The control group was matched to the experimental group and was selected considering age, sex, and religion. The experimental group participated in the memory training program. The memory training program was based on the literature of Fogler & Stern (1994), Wang & Lee (1990), Lee (1991) and Lee (1993). The memory training program was given twice a week for two weeks with each program lasting two hours. Task centered memory self-efficacy was measured using the Memory Self-Efficacy Scale developed by Berry & Dennehey (1989) and Meta Memory was measured by the MIA developed by Dixon et al. (1988) Memory performance was measured by the word list developed by Cho Sung Won (1995) and the face recognition task (Face Recognition Task developed for this study). Data were analyzed by SPSS PC and the results are described below. 1. The experimental group which participated in the Memory Training Program showed higher task centered memory self-efficacy scores as compared to the control group (t=4.354, P=.0001). 2. The experimental group which participated in the Memory Training Program showed higher metamemory scores as compared to the control group (t=4.733, P=.0001). 3. The experimental group which participated in the Memory Training Program showed higher memory performance scores as compared to the control group (t=7.500, P=.0001). The memory performance involved an immediate word recall task, a delayed word recall task, a word recognition task, and the face recognition task. 4. In the experimental group, there was significant correlation between the task centered memory self-efficacy scores and the metamemory scores (r=.382, P=.006), but the correlation between the task centered memory self-efficacy scores and the memory performance scores and between the metamemory scores and the memory performance scores were not significant. The results showed that task centered memory self-efficacy, meta memory and memory performance improved following the Memory Training Program including the memory process, changes in memory with aging, and appropriate use of memory strategies. Memory Training Program is an effective nursing intervention for improving memory in elderly people and, also, in people with complaints of memory loss.

• 말소리와 음성과학
• /
• 제1권2호
• /
• pp.85-94
• /
• 2009

The purpose of this study was to see the effects of pitch and length patterns on spoken word recognition. In Experiment 1, a syllable monitoring task was used to see the effects of pitch and length on the pre-lexical level of spoken word recognition. For both Seoul dialect speakers and Daegu dialect speakers, pitch and length did not affect the syllable detection processes. This result implies that there is little effect of pitch and length in pre-lexical processing. In Experiment 2, a lexical decision task was used to see the effect of pitch and length on the lexical access level of spoken word recognition. In this experiment, word frequency (low and high) as well as pitch and length was manipulated. The results showed that pitch and length information did not play an important role for Seoul dialect speakers, but that it did affect lexical decision processing for Daegu dialect speakers. Pitch and length seem to affect lexical access during the word recognition process of Daegu dialect speakers.

• The Journal of the Acoustical Society of Korea
• /
• 제26권1E호
• /
• pp.21-26
• /
• 2007

In this paper, we propose a modified Viterbi algorithm to compensate for endpoint detection error during the decoding phase of an isolated word recognition task. Since the conventional Viterbi algorithm explores only the search space whose boundaries are fixed to the endpoints of the segmented utterance by the endpoint detector, the recognition performance is highly dependent on the accuracy level of endpoint detection. Inaccurately segmented word boundaries lead directly to recognition error. In order to relax the degradation of recognition accuracy due to endpoint detection error, we describe an unconstrained search of word boundaries and present an algorithm to explore the search space with efficiency. The proposed algorithm was evaluated by performing a variety of simulated endpoint detection error cases on an isolated word recognition task. The proposed algorithm reduced the Word Error Rate (WER) considerably, from 84.4% to 10.6%, while consuming only a little more computation power.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 1999년도 춘계학술발표논문집 논문집
• /
• pp.139-145
• /
• 1999

In current study on 8 college students there was examined modulation of eyeblink (as measured by integrated EMG of m.orbicularis oculi) and skin conductance response (SCR) to an acoustic startle probe (85 dB[A] white noise) by attending to task presented in auditory modality (to memorize words for further recognition) and entire performance of the word recognition test. Both eyeblink magnitude and SCR amplitude and rise time to startle probes were modified (larger magnitude of EMG peak, lower amplitude and shorter rise time of SCR) during attending to task as compared to performance on test. Results are interpreted n terms of modification of electrodermal and eyeblink components of startle and orienting reflexes by task characteristics (passive versus active efforts), attentional demands and aversiveness of experimental situation. However, eyeblink startle response manifested potentiation during attending to task, while SCR demonstrated attenuation. There are discussed implications of startle modulatioas a potentially sensitive probe of situational demands in stress research and also are considered prospects for further studies.

• 한국노년학
• /
• 제36권4호
• /
• pp.1109-1124
• /
• 2016

본 연구에서는 고정관념 위협이 노인의 공간 기억 및 정서인식기능에 미치는 영향을 확인하고, 개인이 지니고 있는 노화에 대한 인식에 따라 고정관념 위협의 효과가 다르게 나타나는지 검증해 보고자 하였다. 이를 위해 자발적으로 연구 참여 의사를 밝힌 60세 이상 노인 17명(남=7)을 대상으로 연구를 진행하였으며, 첫 번째 방문 시 K-WMS-IV와 MMSE를 포함한 기본 인지기능 검사를 실시하고, 자신의 노화에 대한 인식, 노화 불안, 노화에 대한 태도, 연령 정체성 척도에 응답하도록 하였다. 두 번째 방문 시, 실험군의 경우 노화가 인지기능을 저하시킨다는 스크립트를 읽도록 하여 고정관념 위협에 노출시켰으며, 대조군의 경우 중립적인 스크립트를 읽도록 하였다. 고정관념위협을 조작한 이후 공간 작업기억 과제 (콜시 블록 태핑 과제)와 정서 인식 과제 (얼굴표정 정서인식 과제)를 수행하도록 하고, 수행의 정확도를 관찰하였다. 연구 결과, 고정관념 위협에 노출된 노인 군이 그렇지 않은 노인 군에 비해 정서인식 과제에서 유의하게 저조한 수행 정확도 (p<.05)를 보였다. 또한 자신의 노화에 대한 인식과 고정관념 위협 사이에 상호작용 효과가 확인되어(p<.05), 자신의 노화에 대한 인식이 긍정적인 노인들은 고정관념 위협에 노출되더라도 정서인식 과제와 어려운 공간 작업기억 과제에서 대조군과 유사한 수행 정확도를 보인 반면, 자신의 노화에 대한 인식이 부정적인 노인들은 고정관념 위협에 노출되었을 때, 매우 저조한 수행 정확도를 나타내는 양상을 보였다. 따라서 본 연구에서는 고정관념 위협이 노인의 정서인식 기능에 부정적 영향을 미치는 것을 검증하였으며, 자신의 노화에 대한 긍정적 인식이 고정관념 위협으로 인한 인지기능 저하의 보호 요인으로 작용한다는 것을 확인하였다.