Multimodal sentiment analysis utilizes multimodal data such as text,facial expressions and voice to detect people’s attitudes.With the advent of distributed data collection and annotation,we can easily obtain and sha...Multimodal sentiment analysis utilizes multimodal data such as text,facial expressions and voice to detect people’s attitudes.With the advent of distributed data collection and annotation,we can easily obtain and share such multimodal data.However,due to professional discrepancies among annotators and lax quality control,noisy labels might be introduced.Recent research suggests that deep neural networks(DNNs)will overfit noisy labels,leading to the poor performance of the DNNs.To address this challenging problem,we present a Multimodal Robust Meta Learning framework(MRML)for multimodal sentiment analysis to resist noisy labels and correlate distinct modalities simultaneously.Specifically,we propose a two-layer fusion net to deeply fuse different modalities and improve the quality of the multimodal data features for label correction and network training.Besides,a multiple meta-learner(label corrector)strategy is proposed to enhance the label correction approach and prevent models from overfitting to noisy labels.We conducted experiments on three popular multimodal datasets to verify the superiority of ourmethod by comparing it with four baselines.展开更多
Based on MATRIXx, a universal real-time visual distributed simulation system is developed. The system can receive different input data from network or local terminal. Application models in the simulation modules can a...Based on MATRIXx, a universal real-time visual distributed simulation system is developed. The system can receive different input data from network or local terminal. Application models in the simulation modules can automatically get such data to be analyzed and calculated, and then produce real-time simulation control information. Meanwhile, this paper designs relevant simulation components to implement the input and output data, which can guarantee the real-time and universal of the data transmission. Result of the experimental system shows that the real-time performance of the simulation is perfect.展开更多
In the paper,the experience of contradictions and conflicts as a normal feature of social change and evolutionary trends,is discussed.Contradictions normally destabilize a system,causing dilemmas and conflicts.However...In the paper,the experience of contradictions and conflicts as a normal feature of social change and evolutionary trends,is discussed.Contradictions normally destabilize a system,causing dilemmas and conflicts.However,the paper warns against thinking of destabilization as dysfunctional,as complex systems require a high degree of instability to deal with changes in the environment.Dealing with instabilities requires systems,which can produce alarms that signal a need to regain additional options for development and introduce them into the planning process.The paper introduces the concept of a place alarm system,which uses visions and utopias in various forms for alarm purposes.To illustrate this kind of thinking,the paper first presents a model of an alarm system,which demonstrates how we can deal with contradictions by including the temporal dimension in our analyses of place systems and bring broader temporal horizons into consideration.As an example,contradictions increase when we consider the future from the perspective of the present;the present future multiplies contradictions.On the other hand,viewing the present from the future(future present),creates possibilities for goal-directed planning to avoid the problems,which have produced alarm signals.As the paper demonstrates,these two possibilities of what may be called reflexive and utopian temporal modalizations,are not given as alternatives,but mutually imply each other.The paper then presents two case illustrations,which demonstrate how to deal with conflict and contradictions to facilitate collective and goal-directed planning,using the alarm system framework.In both cases,we are witnessing place planning processes,which lack the necessary requisite variety(Ashby,1956)to deal effectively with the environmental complexity and internal conflicts facing the local communities.As indicated in the analyses,the present planning regimes do not promote variety and vitality regarding current place developments.展开更多
A distributed information network with complex network structure always has a challenge of locating fault root causes.In this paper,we propose a novel root cause analysis(RCA)method by random walk on the weighted faul...A distributed information network with complex network structure always has a challenge of locating fault root causes.In this paper,we propose a novel root cause analysis(RCA)method by random walk on the weighted fault propagation graph.Different from other RCA methods,it mines effective features information related to root causes from offline alarms.Combined with the information,online alarms and graph relationship of network structure are used to construct a weighted graph.Thus,this approach does not require operational experience and can be widely applied in different distributed networks.The proposed method can be used in multiple fault location cases.The experiment results show the proposed approach achieves much better performance with 6%higher precision at least for root fault location,compared with three baseline methods.Besides,we explain how the optimal parameter’s value in the random walk algorithm influences RCA results.展开更多
基金supported by STI 2030-Major Projects 2021ZD0200400National Natural Science Foundation of China(62276233 and 62072405)Key Research Project of Zhejiang Province(2023C01048).
文摘Multimodal sentiment analysis utilizes multimodal data such as text,facial expressions and voice to detect people’s attitudes.With the advent of distributed data collection and annotation,we can easily obtain and share such multimodal data.However,due to professional discrepancies among annotators and lax quality control,noisy labels might be introduced.Recent research suggests that deep neural networks(DNNs)will overfit noisy labels,leading to the poor performance of the DNNs.To address this challenging problem,we present a Multimodal Robust Meta Learning framework(MRML)for multimodal sentiment analysis to resist noisy labels and correlate distinct modalities simultaneously.Specifically,we propose a two-layer fusion net to deeply fuse different modalities and improve the quality of the multimodal data features for label correction and network training.Besides,a multiple meta-learner(label corrector)strategy is proposed to enhance the label correction approach and prevent models from overfitting to noisy labels.We conducted experiments on three popular multimodal datasets to verify the superiority of ourmethod by comparing it with four baselines.
文摘Based on MATRIXx, a universal real-time visual distributed simulation system is developed. The system can receive different input data from network or local terminal. Application models in the simulation modules can automatically get such data to be analyzed and calculated, and then produce real-time simulation control information. Meanwhile, this paper designs relevant simulation components to implement the input and output data, which can guarantee the real-time and universal of the data transmission. Result of the experimental system shows that the real-time performance of the simulation is perfect.
文摘In the paper,the experience of contradictions and conflicts as a normal feature of social change and evolutionary trends,is discussed.Contradictions normally destabilize a system,causing dilemmas and conflicts.However,the paper warns against thinking of destabilization as dysfunctional,as complex systems require a high degree of instability to deal with changes in the environment.Dealing with instabilities requires systems,which can produce alarms that signal a need to regain additional options for development and introduce them into the planning process.The paper introduces the concept of a place alarm system,which uses visions and utopias in various forms for alarm purposes.To illustrate this kind of thinking,the paper first presents a model of an alarm system,which demonstrates how we can deal with contradictions by including the temporal dimension in our analyses of place systems and bring broader temporal horizons into consideration.As an example,contradictions increase when we consider the future from the perspective of the present;the present future multiplies contradictions.On the other hand,viewing the present from the future(future present),creates possibilities for goal-directed planning to avoid the problems,which have produced alarm signals.As the paper demonstrates,these two possibilities of what may be called reflexive and utopian temporal modalizations,are not given as alternatives,but mutually imply each other.The paper then presents two case illustrations,which demonstrate how to deal with conflict and contradictions to facilitate collective and goal-directed planning,using the alarm system framework.In both cases,we are witnessing place planning processes,which lack the necessary requisite variety(Ashby,1956)to deal effectively with the environmental complexity and internal conflicts facing the local communities.As indicated in the analyses,the present planning regimes do not promote variety and vitality regarding current place developments.
基金supported by ZTE Industry-University-Institute Cooperation Funds under Grant No.HC-CN-20201120009。
文摘A distributed information network with complex network structure always has a challenge of locating fault root causes.In this paper,we propose a novel root cause analysis(RCA)method by random walk on the weighted fault propagation graph.Different from other RCA methods,it mines effective features information related to root causes from offline alarms.Combined with the information,online alarms and graph relationship of network structure are used to construct a weighted graph.Thus,this approach does not require operational experience and can be widely applied in different distributed networks.The proposed method can be used in multiple fault location cases.The experiment results show the proposed approach achieves much better performance with 6%higher precision at least for root fault location,compared with three baseline methods.Besides,we explain how the optimal parameter’s value in the random walk algorithm influences RCA results.
