• 대한음성학회지:말소리
• /
• 제56호
• /
• pp.185-194
• /
• 2005

In this paper, we present POSSDM (POSTECH Situation-Based Dialogue Manager) for a spoken dialogue system using a new example and situation-based dialogue management technique for effective generation of appropriate system responses. Spoken dialogue system should generate cooperative responses to smoothly control dialogue flow with the users. We introduce a new dialogue management technique incorporating dialogue examples and situation-based rules for EPG (Electronic Program Guide) domain. For the system response inference, we automatically construct and index a dialogue example database from dialogue corpus, and the best dialogue example is retrieved for a proper system response with the query from a dialogue situation including a current user utterance, dialogue act, and discourse history. When dialogue corpus is not enough to cover the domain, we also apply manually constructed situation-based rules mainly for meta-level dialogue management.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2005년도 추계 학술대회 발표논문집
• /
• pp.113-115
• /
• 2005

In this paper, we present POSSDM (POSTECH Situation-Based Dialogue Manager) for a spoken dialogue system using a new example and situation-based dialogue management techniques for effective generation of appropriate system responses. Spoken dialogue system should generate cooperative responses to smoothly control dialogue flow with the users. We introduce a new dialogue management technique incorporating dialogue examples and situation-based rules for EPG (Electronic Program Guide) domain. For the system response inference, we automatically construct and index a dialogue example database from dialogue corpus, and the best dialogue example is retrieved for a proper system response with the query from a dialogue situation including a current user utterance, dialogue act, and discourse history. When dialogue corpus is not enough to cover the domain, we also apply manually constructed situation-based rules mainly for meta-level dialogue management.

• 대한음성학회지:말소리
• /
• 제68권
• /
• pp.75-93
• /
• 2008

An example-based dialogue system tries to utilize many pairs which are stored in a dialogue database. The most important part of the example-based dialogue system is to find the most similar utterance to user's input utterance. Modality, which is characterized as conveying the speaker's involvement in the propositional content of a given utterance, is one of the core sentence features. For example, the sentence "I want to go to school." has a modality of hope. In this paper, we have proposed a modality classification system which can predict sentence modality in order to improve the performance of example-based dialogue systems. We also define a modality tag set for a dialogue system, and validate this tag set using a rule-based modality classification system. Experimental results show that our modality tag set and modality classification system improve the performance of an example-based dialogue system.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제35권1호
• /
• pp.33-40
• /
• 2008

인간은 다른 사람과 대화할 때, 시행착오 과정을 거치면서 상대방에 관한 학습이 일어난다. 본 논문에서는 이런 과정의 강화학습법(Reinforcement Learning)을 이용하여 대화시스템에 적응형 능력의 부여 방법을 제안한다. 적응형 대화 전략이란 대화시스템이 사용자의 대화 처리 습성을 학습하고, 사용자 만족도와 효율성을 높이는 것을 말한다. 강화 학습법을 효율적으로 대화처리 시스템에 적용하기 위하여 대화를 주 대화와 부대화로 나누어 정의하고 사용하였다. 주 대화에서는 전체적인 만족도를, 부 대화에서는 완료 여부, 완료시간, 에러 횟수를 이용해서 시스템의 효율성을 측정하였다. 또한 학습 과정에서의 사용자 편의성을 위하여 시스템 사용 역량에 따라 사용자를 두 그룹으로 분류한 후 해당 그룹의 강화 학습 훈련 정책을 적용하였다. 실험에서는 개인별, 그룹별 강화 학습에 따라 제안한 방법의 성능을 평가하였다.

• 디지털융복합연구
• /
• 제5권1호
• /
• pp.69-76
• /
• 2007

This paper presents a study on the introduction of User Model-Based Dialogue System. Also we present a plan-based Korean dialogue system as a natural language database user interface for product search. The system can be characterized by its support for mixed initiative to give user more control over dialogue, employment of user model to reflect user's preferences, alternative solution suggestion if there is no product matched exactly to user's requirements, handling circumlocution which frequently occurs in dialogues. The user modeling shell system BGP-MS is adapted for the system. The system provides for a user-friendly database user interface by managing dialogue intelligently. By its implementation and test, it has been shown that the user model-based dialogue system can be utilized effectively for product search.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2009년도 학술대회
• /
• pp.257-262
• /
• 2009

본 논문은 예제기반 대화시스템의 대화 로그를 분석하여 새로운 대화예제를 자동으로 확충하는 방법에 대해서 논한다. 전통적인 예제기반 대화 시스템에서는 사용자와 챗봇(chatbot)간의 대화예제를 수작업으로 구축 하기 때문에 많은 시간과 비용이 소요되는 단점을 가지고 있다. 대화 로그는 대화 시스템의 서비스가 지속되는 한 끊임없이 생성이 되며, 시간이 지날수록 풍부하고 다양한 대화예제를 획득할 수 있다는 장점을 갖는다. 본 논문은 대화로그로부터 자연스러운 발화쌍을 분류하고, 분류된 발화쌍을 대화예제로 추가하는 방법을 제안한다. 최대 엔트로피모형을 사용한 실험의 결과, 자연스러운 발화를 결정하기 위해서는 발화의 어휘, 품사, 양태와 같은 자질이 유용하게 사용될 수 있음을 관찰하였고, 수작업 구축에만 의존하던 대화예제가 대화 로그를 사용하여 자동으로 확충될 수 있음을 확인하였다.

• 한국디지털정책학회:학술대회논문집
• /
• 한국디지털정책학회 2007년도 춘계학술대회
• /
• pp.287-296
• /
• 2007

In this paper we present a plan-based Korean dialogue system as a natural language database user interface for product search. The system can be characterized by its support for mixed initiative to give user more control over dialogue, employment of user model to reflect user' spreferences, alternative solution suggestion if there is no product matched exactly to user's requirements, handling circumlocution which frequently occurs in dialogues. The user model ing shell system BGP-MS is adapted for the system. The system provides for a user-friendly database user interface bymanaging dialogue intelligently. By its implementation and test it has been shown that the user model-based dialogue system can be utilized effectively for product search.

• 전자공학회논문지B
• /
• 제33B권3호
• /
• pp.159-168
• /
• 1996

In this paper, the natural language dialogue understanding sytem, based on discourse information and plan recognition, is designed and implemented. The system needs to analyze the user's input utterance and acquire the discoruse information to perform plan recognition and facilitate cooperative response. This paper proposes the mehtod of controlling a dialogue, based on the algorithm for extracting the discourse information. When the discourse information for dialogue understanding is extracted, the information-based value in feature structure that is obtained form korean parser is used. And the system makes use of the structure. Thus it can offer the response that the user wants to take, and let the dialogue to study in utterance level and enhance the efficiency of dialogue understanding. In this paper, we apply the system to the hotel reservation domain and show the mehtod of using the discoruse information to control the dialogue.

• 인지과학
• /
• 제19권2호
• /
• pp.107-123
• /
• 2008

대화 시스템에 관한 기존 연구는 실용시스템을 위한 연구와 대화 현상 이해에 중심을 둔 연구로 나눌 수 있다. 전자는 실제 환경에서 강인한 성능을 유지하는 것에 초점을 두고 있다. 하지만 매우 제한적인 상황에서만 성능을 보장할 수 있으며 다양한 형태의 대화를 적용하기 어렵다. 후자의 연구에서는 대표적으로 계획기반 모델(plan-based model)이 제안되었는데, 이 모델은 복잡한 대화 구조를 모델링(modeling)할 수 있으며, 다양한 현상에서의 사용자 목적 추론이 가능하다. 하지만 계획기반 모델은 초기 설계가 어렵고 실용적인 대화 시스템 구축에 있어서 시스템 응답을 생성하기 위한 상호작용 모델로의 확장이 매우 어렵기 때문에 실용시스템으로의 적용이 쉽지 않다. 본 연구는 이러한 계획기반 모델의 단점을 보완하고 실용적인 대화 시스템을 구축하기 위하여, 시스템 응답을 위한 확인 대화 전략과 담화스택(discourse stack)을 계획기반 대화 모델에 적용함으로써 효율적인 응답을 생성할 수 있는 새로운 기법을 제안한다.

• 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.