Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255158
Jinguang Chen, Tingting He, Zhuoming Gui, Fang Li
Research on sentence compression has been undergoing for many years in other languages, especially in English, but research on Chinese sentence compression is rarely found. In this paper, we describe an efficient probabilistic and syntactic approach to Chinese sentence compression. We introduce the classical noisy-channel approach into Chinese sentence compression and improve it in many ways. Since there is no parallel training corpus in Chinese, we use the unsupervised learning method. This paper also presents a novel bottom-up optimizing algorithm which considers both bigram and syntactic probabilities for generating candidate compressed sentences. We evaluate results against manual compressions and a simple baseline. The experiments show the effectiveness of the proposed approach.
{"title":"Probabilistic unsupervised Chinese sentence compression","authors":"Jinguang Chen, Tingting He, Zhuoming Gui, Fang Li","doi":"10.1109/GRC.2009.5255158","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255158","url":null,"abstract":"Research on sentence compression has been undergoing for many years in other languages, especially in English, but research on Chinese sentence compression is rarely found. In this paper, we describe an efficient probabilistic and syntactic approach to Chinese sentence compression. We introduce the classical noisy-channel approach into Chinese sentence compression and improve it in many ways. Since there is no parallel training corpus in Chinese, we use the unsupervised learning method. This paper also presents a novel bottom-up optimizing algorithm which considers both bigram and syntactic probabilities for generating candidate compressed sentences. We evaluate results against manual compressions and a simple baseline. The experiments show the effectiveness of the proposed approach.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"84 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115233841","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255140
Jia-li Feng
A new kind of Computer, called Attribute Grid Computer based on Qualitative Mapping is presented in this paper, It is shown that a series of intelligent methods, such as Production System, Artificial Neural Network, and Support Vector Machine can be fused in the framework of qualitative criterion transformation of qualitative mapping and can be implemented by attribute grid computer. And some examples of application in pattern recognition are given too.
{"title":"Attribute Grid Computer based on Qualitative Mapping and its application in pattern Recognition","authors":"Jia-li Feng","doi":"10.1109/GRC.2009.5255140","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255140","url":null,"abstract":"A new kind of Computer, called Attribute Grid Computer based on Qualitative Mapping is presented in this paper, It is shown that a series of intelligent methods, such as Production System, Artificial Neural Network, and Support Vector Machine can be fused in the framework of qualitative criterion transformation of qualitative mapping and can be implemented by attribute grid computer. And some examples of application in pattern recognition are given too.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114324629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255051
Jianwei Xiang, Xia Ke
Analysis of how to extract medical diagnosis rules from medical cases. Based on the rough set theory, a way of acquiring knowledge is introduced. Using this theory, we analyze the data, propose some possible rules and reveal an optimized probability formula. The steps of implementation, which includes the continual information discrimination system, information reduction system, decision acquirement rules, decision model generation, etc., are explained through case study. In the end, the whole process of knowledge acquirement is discussed, which can effectively solve the choke point problem of acquiring knowledge in the expert system. At the same time, it also provides a new way to solve the application of artificial intelligence technology in the field of medicinal diagnosis.
{"title":"A novel extracting medical diagnosis rules based on rough sets","authors":"Jianwei Xiang, Xia Ke","doi":"10.1109/GRC.2009.5255051","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255051","url":null,"abstract":"Analysis of how to extract medical diagnosis rules from medical cases. Based on the rough set theory, a way of acquiring knowledge is introduced. Using this theory, we analyze the data, propose some possible rules and reveal an optimized probability formula. The steps of implementation, which includes the continual information discrimination system, information reduction system, decision acquirement rules, decision model generation, etc., are explained through case study. In the end, the whole process of knowledge acquirement is discussed, which can effectively solve the choke point problem of acquiring knowledge in the expert system. At the same time, it also provides a new way to solve the application of artificial intelligence technology in the field of medicinal diagnosis.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"122 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115249097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255032
Jun Yang, Zhangyan Xu
The key of attribute reduction based on rough set is find the core attributes. Most existing works are mainly based on Hu's discernibility matrix. Till now, there are three kinds of core attributes: Hu's core based on discernibility matrix (denoted by Core1(C)), core based on positive region (denoted by Core2(C)), and core based on information entropy (denoted by Core3(C)). Some researchers have been pointed out that these three kinds of cores are not equivalent to each other. Based on the above three kinds of core attributes, we at first propose three kinds of simplified discernibility matrices and their corresponding cores, which are denoted by SDCore1(C), SDCore2(C), and SDCore3(C) respectively. And then it is proved that Core1(C)=SDCore1(C), Core2(C)= SDCore2(C), and Core3(C)=SDCore3(C). Finally, based on three proposed simplified discernibility matrices and their corresponding cores, it is proved that Core2(C)⊆Core3(C)⊆Core1(C).
{"title":"Different core attributes's comparison and analysis","authors":"Jun Yang, Zhangyan Xu","doi":"10.1109/GRC.2009.5255032","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255032","url":null,"abstract":"The key of attribute reduction based on rough set is find the core attributes. Most existing works are mainly based on Hu's discernibility matrix. Till now, there are three kinds of core attributes: Hu's core based on discernibility matrix (denoted by Core1(C)), core based on positive region (denoted by Core2(C)), and core based on information entropy (denoted by Core3(C)). Some researchers have been pointed out that these three kinds of cores are not equivalent to each other. Based on the above three kinds of core attributes, we at first propose three kinds of simplified discernibility matrices and their corresponding cores, which are denoted by SDCore1(C), SDCore2(C), and SDCore3(C) respectively. And then it is proved that Core1(C)=SDCore1(C), Core2(C)= SDCore2(C), and Core3(C)=SDCore3(C). Finally, based on three proposed simplified discernibility matrices and their corresponding cores, it is proved that Core2(C)⊆Core3(C)⊆Core1(C).","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"327 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123314086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255082
Ji Ma
Genetic algorithms have been applied in various application domains and research fields related to biology, chemistry, especially computer science and engineering. In this paper, we will discuss the applications of generic algorithms in project scheduling. The problem is described, the algorithm is outlined, and the strengths and weaknesses are compared. Finally, the future trends in this direction are predicted.
{"title":"Project scheduling based on genetic algorithm","authors":"Ji Ma","doi":"10.1109/GRC.2009.5255082","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255082","url":null,"abstract":"Genetic algorithms have been applied in various application domains and research fields related to biology, chemistry, especially computer science and engineering. In this paper, we will discuss the applications of generic algorithms in project scheduling. The problem is described, the algorithm is outlined, and the strengths and weaknesses are compared. Finally, the future trends in this direction are predicted.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"54 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124798029","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255080
S. Miyamoto, Mitsuaki Yamazaki, Wataru Hashimoto
This paper discusses a method of semi-supervised fuzzy clustering with target clusters. The method uses two kinds of additional terms to ordinary fuzzy c-means objective function. One term consists of the sum of squared differences between the target cluster memberships and the membership of the solution, whereas second term has the sum of absolute differences of those memberships. While the former has a closed formula for the membership solution, the second requires a complicated algorithm. However, numerical example show that the latter method of the absolute differences works better.
{"title":"Fuzzy semi-supervised clustering with target clusters using different additional terms","authors":"S. Miyamoto, Mitsuaki Yamazaki, Wataru Hashimoto","doi":"10.1109/GRC.2009.5255080","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255080","url":null,"abstract":"This paper discusses a method of semi-supervised fuzzy clustering with target clusters. The method uses two kinds of additional terms to ordinary fuzzy c-means objective function. One term consists of the sum of squared differences between the target cluster memberships and the membership of the solution, whereas second term has the sum of absolute differences of those memberships. While the former has a closed formula for the membership solution, the second requires a complicated algorithm. However, numerical example show that the latter method of the absolute differences works better.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124854330","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255030
Qingyun Yang, Chunjie Wang, Changsheng Zhang
Task Assignment Problems (TAPs) in distributed computer system are general NP-hard and usually modeled as integer programming discrete problems. Many algorithms are proposed to resolve those problems. Discrete particle swarm algorithm (DPS) is a newly developed method to solve constraint satisfaction problem (CSP) which has advantage on search capacity and can find more solutions. We proposed an improved DPS to solve TAP in this paper. DPS has a special operator namely coefficient multiplying speed, which is designed for CSP but does not exist in other discrete problems. Thus we redefined a coefficient multiplying speed operator with probability selection. We analyzed the speed and position updating formula then we derived a refined position updating formula. Several experiments are carried out to test our DPS. Experimental results show that our algorithm has more efficient search capacity, higher success rate, less running time and more robust.
{"title":"An efficient discrete particle swarm algorithm for Task Assignment Problems","authors":"Qingyun Yang, Chunjie Wang, Changsheng Zhang","doi":"10.1109/GRC.2009.5255030","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255030","url":null,"abstract":"Task Assignment Problems (TAPs) in distributed computer system are general NP-hard and usually modeled as integer programming discrete problems. Many algorithms are proposed to resolve those problems. Discrete particle swarm algorithm (DPS) is a newly developed method to solve constraint satisfaction problem (CSP) which has advantage on search capacity and can find more solutions. We proposed an improved DPS to solve TAP in this paper. DPS has a special operator namely coefficient multiplying speed, which is designed for CSP but does not exist in other discrete problems. Thus we redefined a coefficient multiplying speed operator with probability selection. We analyzed the speed and position updating formula then we derived a refined position updating formula. Several experiments are carried out to test our DPS. Experimental results show that our algorithm has more efficient search capacity, higher success rate, less running time and more robust.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122669003","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255163
Kunrong Chen, Fen Lin, Qing Tan, Zhongzhi Shi
A basic problem of intelligent systems is choosing adaptive action to perform in a non-stationary environment. Due to the combinatorial complexity of actions, agent cannot possibly consider every option available to it at every instant in time. It needs to find good policies that dictate optimum actions to perform in each situation. This paper proposes an algorithm, called UQ-learning, to better solve action selection problem by using reinforcement learning and utility function. Reinforcement learning can provide the information of environment and utility function is used to balance Exploration-Exploitation dilemma. We implement our method with maze navigation tasks in a non-stationary environment. The results of simulated experiments show that utility-based reinforcement learning approach is more effective and efficient compared with Q-learning and Recency-Based Exploration.
{"title":"Adaptive action selection using utility-based reinforcement learning","authors":"Kunrong Chen, Fen Lin, Qing Tan, Zhongzhi Shi","doi":"10.1109/GRC.2009.5255163","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255163","url":null,"abstract":"A basic problem of intelligent systems is choosing adaptive action to perform in a non-stationary environment. Due to the combinatorial complexity of actions, agent cannot possibly consider every option available to it at every instant in time. It needs to find good policies that dictate optimum actions to perform in each situation. This paper proposes an algorithm, called UQ-learning, to better solve action selection problem by using reinforcement learning and utility function. Reinforcement learning can provide the information of environment and utility function is used to balance Exploration-Exploitation dilemma. We implement our method with maze navigation tasks in a non-stationary environment. The results of simulated experiments show that utility-based reinforcement learning approach is more effective and efficient compared with Q-learning and Recency-Based Exploration.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122724932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255159
Hongmei Chen, Tianrui Li, Weibin Liu, Weili Zou
In rough set theory (RST), upper and lower approximations of a concept will change dynamically while the information system varies over time. How to update approximations based on the original approximations' information is an important problem since it may improve the efficiency of knowledge discovery. This paper focuses on the approach for dynamically updating approximations when attribute values coarsening or refining. The definitions of attribute values coarsening and refining in information systems are introduced. The properties for dynamic maintenance of upper and lower approximations while attribute values coarsen and refine are presented. Finally, the principle of coarsening or refining of the multi-granularity attribute values is analyzed.
{"title":"Research on the approach of dynamically maintenance of approximations in rough set theory while attribute values coarsening and refining","authors":"Hongmei Chen, Tianrui Li, Weibin Liu, Weili Zou","doi":"10.1109/GRC.2009.5255159","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255159","url":null,"abstract":"In rough set theory (RST), upper and lower approximations of a concept will change dynamically while the information system varies over time. How to update approximations based on the original approximations' information is an important problem since it may improve the efficiency of knowledge discovery. This paper focuses on the approach for dynamically updating approximations when attribute values coarsening or refining. The definitions of attribute values coarsening and refining in information systems are introduced. The properties for dynamic maintenance of upper and lower approximations while attribute values coarsen and refine are presented. Finally, the principle of coarsening or refining of the multi-granularity attribute values is analyzed.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"65 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126258938","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2009-09-22DOI: 10.1109/GRC.2009.5255035
Qing Yang, Wei Chen, Bin Wen
For expressing the fuzziness and uncertainty of domain knowledge, realizing the semantic retrieval of fuzzy information, this paper produces an extended fuzzy ontology model and proposes a kind of semantic query expansion technology which can implement semantic information query based on the property values and the relationships of fuzzy concepts. The extended fuzzy ontology provides appropriate support for Learning Evaluation. To access the effect of the proposed model, many experiments have been given for the performance evaluation. The results show that this system can improve retrieval accuracy and promote intelligent semantic query.
{"title":"Fuzzy ontology generation model using fuzzy clustering for learning evaluation","authors":"Qing Yang, Wei Chen, Bin Wen","doi":"10.1109/GRC.2009.5255035","DOIUrl":"https://doi.org/10.1109/GRC.2009.5255035","url":null,"abstract":"For expressing the fuzziness and uncertainty of domain knowledge, realizing the semantic retrieval of fuzzy information, this paper produces an extended fuzzy ontology model and proposes a kind of semantic query expansion technology which can implement semantic information query based on the property values and the relationships of fuzzy concepts. The extended fuzzy ontology provides appropriate support for Learning Evaluation. To access the effect of the proposed model, many experiments have been given for the performance evaluation. The results show that this system can improve retrieval accuracy and promote intelligent semantic query.","PeriodicalId":388774,"journal":{"name":"2009 IEEE International Conference on Granular Computing","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126272126","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：