In order to overcome defects of the classical hidden Markov model (HMM), Markov family model (MFM), a new statistical model was proposed. Markov family model was applied to speech recognition and natural language proc...In order to overcome defects of the classical hidden Markov model (HMM), Markov family model (MFM), a new statistical model was proposed. Markov family model was applied to speech recognition and natural language processing. The speaker independently continuous speech recognition experiments and the part-of-speech tagging experiments show that Markov family model has higher performance than hidden Markov model. The precision is enhanced from 94.642% to 96.214% in the part-of-speech tagging experiments, and the work rate is reduced by 11.9% in the speech recognition experiments with respect to HMM baseline system.展开更多
Natural language parsing is a task of great importance and extreme difficulty. In this paper, we present a full Chinese parsing system based on a two-stage approach. Rather than identifying all phrases by a uniform mo...Natural language parsing is a task of great importance and extreme difficulty. In this paper, we present a full Chinese parsing system based on a two-stage approach. Rather than identifying all phrases by a uniform model, we utilize a divide and conquer strategy. We propose an effective and fast method based on Markov model to identify the base phrases. Then we make the first attempt to extend one of the best English parsing models i.e. the head-driven model to recognize Chinese complex phrases. Our two-stage approach is superior to the uniform approach in two aspects. First, it creates synergy between the Markov model and the head-driven model. Second, it reduces the complexity of full Chinese parsing and makes the parsing system space and time efficient. We evaluate our approach in PARSEVAL measures on the open test set, the parsing system performances at 87.53% precision, 87.95% recall.展开更多
Integrating multiple systems into one has become an important trend in Process Systems Engineering research field since there is strong demand from the modern industries. In this study, a stage-wise superstructurebase...Integrating multiple systems into one has become an important trend in Process Systems Engineering research field since there is strong demand from the modern industries. In this study, a stage-wise superstructurebased method is proposed to synthesize a combined mass and heat exchange network(CM&HEN) which has two parts as the mass exchange network(MEN) and heat exchange network(HEN) involved. To express the possible heat exchange requirements resulted from mass exchange operations, a so called "indistinct HEN superstructure(IHS)", which can contain the all potential matches between streams, is constructed at first. Then, a non-linear programming(NLP) mathematical model is established for the simultaneous synthesis and optimization of networks. Therein, the interaction between mass exchange and heat exchange is modeling formulated.The NLP model has later been examined using an example from literature, and the effectiveness of the proposed method has been demonstrated with the results.展开更多
基金Project(60763001)supported by the National Natural Science Foundation of ChinaProjects(2009GZS0027,2010GZS0072)supported by the Natural Science Foundation of Jiangxi Province,China
文摘In order to overcome defects of the classical hidden Markov model (HMM), Markov family model (MFM), a new statistical model was proposed. Markov family model was applied to speech recognition and natural language processing. The speaker independently continuous speech recognition experiments and the part-of-speech tagging experiments show that Markov family model has higher performance than hidden Markov model. The precision is enhanced from 94.642% to 96.214% in the part-of-speech tagging experiments, and the work rate is reduced by 11.9% in the speech recognition experiments with respect to HMM baseline system.
基金国家高技术研究发展计划(863计划)，the National Natural Science Foundation of China
文摘Natural language parsing is a task of great importance and extreme difficulty. In this paper, we present a full Chinese parsing system based on a two-stage approach. Rather than identifying all phrases by a uniform model, we utilize a divide and conquer strategy. We propose an effective and fast method based on Markov model to identify the base phrases. Then we make the first attempt to extend one of the best English parsing models i.e. the head-driven model to recognize Chinese complex phrases. Our two-stage approach is superior to the uniform approach in two aspects. First, it creates synergy between the Markov model and the head-driven model. Second, it reduces the complexity of full Chinese parsing and makes the parsing system space and time efficient. We evaluate our approach in PARSEVAL measures on the open test set, the parsing system performances at 87.53% precision, 87.95% recall.
基金Supported by the Fundamental Research Funds for the Central Universities of China(DUT14RC(3)046)China Postdoctoral Science Foundation(2014M551091)the National Natural Science Foundation of China(21406026)
文摘Integrating multiple systems into one has become an important trend in Process Systems Engineering research field since there is strong demand from the modern industries. In this study, a stage-wise superstructurebased method is proposed to synthesize a combined mass and heat exchange network(CM&HEN) which has two parts as the mass exchange network(MEN) and heat exchange network(HEN) involved. To express the possible heat exchange requirements resulted from mass exchange operations, a so called "indistinct HEN superstructure(IHS)", which can contain the all potential matches between streams, is constructed at first. Then, a non-linear programming(NLP) mathematical model is established for the simultaneous synthesis and optimization of networks. Therein, the interaction between mass exchange and heat exchange is modeling formulated.The NLP model has later been examined using an example from literature, and the effectiveness of the proposed method has been demonstrated with the results.
