Consistency identification in task-oriented dialogue(CI-ToD)can prevent inconsistent dialogue response generation,which has recently emerged as an important and growing research area.This paper takes the first step to...Consistency identification in task-oriented dialogue(CI-ToD)can prevent inconsistent dialogue response generation,which has recently emerged as an important and growing research area.This paper takes the first step to explore a pre-training paradigm for CI-ToD.Nevertheless,pre-training for CI-ToD is non-trivial because it requires a large amount of multi-turn KB-grounded dialogues,which are extremely hard to collect.To alleviate the data scarcity problem for pre-training,we introduce a modularized pre-training framework(MPFToD),which is capable of utilizing large amounts of KB-free dialogues.Specifically,such modularization allows us to decouple CI-ToD into three sub-modules and propose three pre-training tasks including(i)query response matching pre-training;(ii)dialogue history consistent identification pre-training;and(iii)KB mask language modeling to enhance different abilities of CI-ToD model.As different sub-tasks are solved separately,MPFToD can learn from large amounts of KB-free dialogues for different modules,which are much easier to obtain.Results on the CI-ToD benchmark show that MPFToD pushes the state-of-the-art performance from 56.3%to 61.0%.Furthermore,we show its transferability with promising performance on other downstream tasks(i.e.,dialog act recognition,sentiment classification and table fact checking).展开更多
A steady increase in consumer demands, and severe constraints from both a somewhat damaged environment and newly installed government policies, require today's product design and development to be faster and more...A steady increase in consumer demands, and severe constraints from both a somewhat damaged environment and newly installed government policies, require today's product design and development to be faster and more efficient than ever before, yet utilizing even fewer resources. New holistic approaches, such as total product life cycle modeling which embraces all aspects of a product's life cycle, are current attempts to solve these problems. Within the field of product design and modeling, feature technology has proved to be one very promising solution component. Owing to the tremendous increase in information technology, to transfer from low level data processing towards knowledge modeling and information processing is about to bring a change in almost every computerized application. From this viewpoint, current problems of both feature frameworks and feature systems are analyzed in respect to static and dynamic consistency breakdowns. The analysis ranges from early stages of designing (feature) concepts to final system implementation and application. Por the first time, an integrated view is given oil approaches, solutions and practical experience, with feature concepts and structures, providing both a feature framework and its implementation with sufficient system architecture and computational power to master a fair number of known consistency breakdowns, while providing for robust contexts for feature semantics and integrated models. Within today's heavy use of information technology these are pre-requisites if the full potential of feature technology is to be successfully translated into practice.展开更多
为进行有效、规范的体系结构构建,现有体系结构大多采用美国国防部体系结构框架(Department of Defense Architecture Framework,DoDAF)作为指导,针对采用该框架构建的体系结构的数据一致性验证问题,提出了一种基于国防部体系结构框架...为进行有效、规范的体系结构构建,现有体系结构大多采用美国国防部体系结构框架(Department of Defense Architecture Framework,DoDAF)作为指导,针对采用该框架构建的体系结构的数据一致性验证问题,提出了一种基于国防部体系结构框架元模型(DoDAF meta-model,DM2)的验证方法。该方法直接从DM2中提取数据的一致性原则,而后利用查找出来的实体数据构建验证准则或者验证矩阵,进而完成体系结构数据一致性的判断。最后,通过一个案例进行一致性验证步骤说明和可行性分析,验证结果充分说明了该方法的有效性。展开更多
针对7×24小时运行的系统,不能停止系统运行进行更新或维护的问题,根据OSGi(Open Service Gateway Initiative)框架的特点,在分析已有的服务动态更新方法基础上,克服已有策略不能保证服务更新前后的相互一致性和服务状态的一致性,...针对7×24小时运行的系统,不能停止系统运行进行更新或维护的问题,根据OSGi(Open Service Gateway Initiative)框架的特点,在分析已有的服务动态更新方法基础上,克服已有策略不能保证服务更新前后的相互一致性和服务状态的一致性,以及更新策略实现过程中的编程模型不佳问题,提出基于发布/订阅模型的服务动态更新策略。以静止状态理论为依据,利用Petri网形式化描述服务动态更新过程,从理论上验证该策略的有效性;同时,以Event Admin Service作为策略的实现工具,具有良好的编程模型,从而实现OSGi平台上的动态演化,实验结果表明该策略具有可行性。展开更多
基金supported by the National Natural Science Foundation of China(NSFC)(Grant Nos.62306342,62176076)the Excellent Young Scientists Fund in Hunan Province(2024JJ4070)+3 种基金supported by the Natural Science Foundation of Guangdong(2023A1515012922)Shenzhen Foundational Research Funding(JCYJ20220818102415032)The Major Key Project of PCL(PCL2023A09)Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies(2022B1212010005k).
文摘Consistency identification in task-oriented dialogue(CI-ToD)can prevent inconsistent dialogue response generation,which has recently emerged as an important and growing research area.This paper takes the first step to explore a pre-training paradigm for CI-ToD.Nevertheless,pre-training for CI-ToD is non-trivial because it requires a large amount of multi-turn KB-grounded dialogues,which are extremely hard to collect.To alleviate the data scarcity problem for pre-training,we introduce a modularized pre-training framework(MPFToD),which is capable of utilizing large amounts of KB-free dialogues.Specifically,such modularization allows us to decouple CI-ToD into three sub-modules and propose three pre-training tasks including(i)query response matching pre-training;(ii)dialogue history consistent identification pre-training;and(iii)KB mask language modeling to enhance different abilities of CI-ToD model.As different sub-tasks are solved separately,MPFToD can learn from large amounts of KB-free dialogues for different modules,which are much easier to obtain.Results on the CI-ToD benchmark show that MPFToD pushes the state-of-the-art performance from 56.3%to 61.0%.Furthermore,we show its transferability with promising performance on other downstream tasks(i.e.,dialog act recognition,sentiment classification and table fact checking).
文摘A steady increase in consumer demands, and severe constraints from both a somewhat damaged environment and newly installed government policies, require today's product design and development to be faster and more efficient than ever before, yet utilizing even fewer resources. New holistic approaches, such as total product life cycle modeling which embraces all aspects of a product's life cycle, are current attempts to solve these problems. Within the field of product design and modeling, feature technology has proved to be one very promising solution component. Owing to the tremendous increase in information technology, to transfer from low level data processing towards knowledge modeling and information processing is about to bring a change in almost every computerized application. From this viewpoint, current problems of both feature frameworks and feature systems are analyzed in respect to static and dynamic consistency breakdowns. The analysis ranges from early stages of designing (feature) concepts to final system implementation and application. Por the first time, an integrated view is given oil approaches, solutions and practical experience, with feature concepts and structures, providing both a feature framework and its implementation with sufficient system architecture and computational power to master a fair number of known consistency breakdowns, while providing for robust contexts for feature semantics and integrated models. Within today's heavy use of information technology these are pre-requisites if the full potential of feature technology is to be successfully translated into practice.
