基于FCA与统计学习的本体生成技术研究
|
摘要
20世纪90年代以来,信息科学的发展面临着种种新的难题,主要包括知识表示、信息组织、软件利用等。特别是由于因特网的快速发展,如何组织、管理和维护大量信息并为用户提供有效的服务也就成为一项重要而迫切的研究任务。为适应这些要求,哲学领域的本体被引入信息科学领域,由于国内外众多学者的研究,已经发展成为能在语义和知识层次上描述信息系统的概念模型建模工具,并在信息检索、信息抽取、异构信息系统的互操作和集成、语义Web等领域得到了广泛应用。
本文主要围绕本体生成技术及本体的优化方法进行展开。论文在分析当前经典本体生成方法的基础上,提出了基于FCA与统计学习相结合的本体生成算法,提出了在应用FCA相关理论生成概念格时的四个统计量,构造了基于FCA与统计学习相结合生成的概念格到本体的映射过程;基于FCA中概念格的现有理论,特别是继承概念格理论,得出有关冗余概念的推论和优化概念格的定义,从而形成了本体优化算法的基础;论文还对本体迁移学习进行了研究。通过以上本体生成及优化过程,能够得到比较理想的本体结构。最后本文给出了本体生成的框架模型和可视化本体生成平台。
通过本文提出的本体生成技术与方法,能够准确地提取领域知识,建立结构清晰、重点突出的领域知识本体,对于有效组织、管理和维护大量信息,有效地从大量信息中提取关键知识具有重要意义。
Since the 1990s, the development of information science is facing many new problems, mainly, such as knowledge representation, information organization, and software utilization. Particularly because of the rapid development of the Internet, how to organize, manage and maintain a amount of information and provide users with efficient service has become an important and urgent research content. To meet these requirements, ontology, a concept from the field of philosophy, is introduced to the field of information science, due to a number of researches, it has become a concept modeling tools which can describe information systems on the semantic and knowledge levels, and has been widely used in information retrieval, information extraction, heterogeneous information systems interoperability and integration, Semantic Web and other applications.
This paper mainly focuses on generating technologies of ontology construction and ontology optimization method. Paper represents a detailed analysis of the current classical ontology generation methods, with analyzing their strengths and weakness, paper gives the ontology generation algorithms base on the combination of the proposed FCA and statistical learning. Proposes four statistics in the construction of concept lattice and construct mapping process from concept lattice to ontology. In addition, base on the existing theory of FCA concept lattice, in particular, inheritance concept lattice theory, paper draws inferences about the concept of redundancy and the definition of optimizing concept lattice, thus forming the basis of ontology optimization algorithm, and gives a research of similar ontology conducting. Through the above process Ontology, a more ideal ontology structure can be got with ontology generation and optimization. Finally, this paper gives a framework of generating ontology and generates the interface for visualization of ontology construction.
Through the ontology constructing techniques and methods of this paper, which can accurately extract the domain knowledge, and establish a clear and focused domain knowledge ontology, which has great significance to effectively organize, manage, maintain a large amount of information and extract critical knowledge from a large number of information.
本文主要围绕本体生成技术及本体的优化方法进行展开。论文在分析当前经典本体生成方法的基础上,提出了基于FCA与统计学习相结合的本体生成算法,提出了在应用FCA相关理论生成概念格时的四个统计量,构造了基于FCA与统计学习相结合生成的概念格到本体的映射过程;基于FCA中概念格的现有理论,特别是继承概念格理论,得出有关冗余概念的推论和优化概念格的定义,从而形成了本体优化算法的基础;论文还对本体迁移学习进行了研究。通过以上本体生成及优化过程,能够得到比较理想的本体结构。最后本文给出了本体生成的框架模型和可视化本体生成平台。
通过本文提出的本体生成技术与方法,能够准确地提取领域知识,建立结构清晰、重点突出的领域知识本体,对于有效组织、管理和维护大量信息,有效地从大量信息中提取关键知识具有重要意义。
Since the 1990s, the development of information science is facing many new problems, mainly, such as knowledge representation, information organization, and software utilization. Particularly because of the rapid development of the Internet, how to organize, manage and maintain a amount of information and provide users with efficient service has become an important and urgent research content. To meet these requirements, ontology, a concept from the field of philosophy, is introduced to the field of information science, due to a number of researches, it has become a concept modeling tools which can describe information systems on the semantic and knowledge levels, and has been widely used in information retrieval, information extraction, heterogeneous information systems interoperability and integration, Semantic Web and other applications.
This paper mainly focuses on generating technologies of ontology construction and ontology optimization method. Paper represents a detailed analysis of the current classical ontology generation methods, with analyzing their strengths and weakness, paper gives the ontology generation algorithms base on the combination of the proposed FCA and statistical learning. Proposes four statistics in the construction of concept lattice and construct mapping process from concept lattice to ontology. In addition, base on the existing theory of FCA concept lattice, in particular, inheritance concept lattice theory, paper draws inferences about the concept of redundancy and the definition of optimizing concept lattice, thus forming the basis of ontology optimization algorithm, and gives a research of similar ontology conducting. Through the above process Ontology, a more ideal ontology structure can be got with ontology generation and optimization. Finally, this paper gives a framework of generating ontology and generates the interface for visualization of ontology construction.
Through the ontology constructing techniques and methods of this paper, which can accurately extract the domain knowledge, and establish a clear and focused domain knowledge ontology, which has great significance to effectively organize, manage, maintain a large amount of information and extract critical knowledge from a large number of information.
引文
[1]陆建江,张来非,苗壮等.语义网原理与技术[M],科学出版社.67-68.
[2]邓志鸿,唐世渭,张铭等.Ontology研究综述[J].北京大学学报:自然科学版, 2002(5):730-738.
[3]陈静,朱巧明,贡正仙.基于Ontology的信息抽取研究综述[J].计算机技术与发展, 2007(10):84-86.
[4] Nguyen N, Matsumoto M.Ontology-Based Context Management Agent for Vertical Handoff Use Fuzzy Logic Decision in Heterogeneous Network [C]. Proceedings of the First International Conference on Complex, Intelligent and Software Intensive Systems, 2006: 215-220.
[5] Li Jian, Jin Beihong. Query Division and Reformulation in Ontology-Based Heterogeneous Information Integration[C]. Proceedings of the15th International Conference on Computing, 2006:186-196.
[6]MaedcheA, Staab S. Ontology learning for the semantic web [J]. In: IEEE Intelligent Systems, special Issue on the Semantic Web.
[7] Gupta K M, Aha D W, Marsh E, et al. An architecture for engineering sublanguage WordNets [J]. In: Proceedings of the First International Conference on Global WordNet.
[8] Maedche A, Volz R. The text-to-onto ontology extraction and maintance environment [J]. In: Proceedings of the ICDM Workshop on Integrating Data Mining and Knowledge Management. California, USA.
[9] Khan L, Luo F. 2002.Ontology construction of information selection [J]. In: Proceedings of the Fourteenth IEEE International Conference on Tools with Artificial Intelligence. Washington D.C. USA.
[10] Kietz J, Maedche A, Volz R. 2000. A method for semi-automatic ontology acquisition from a corporate intranet[J]. In: Proceedings of EKAW'00 Workshop on Ontologies and Texts.
[11] Hahn U,Marko K G..2001. Joint knowledge capture for grammars and ontologies [J].In:Proceedings of the First International Conference on Knowledge Capture K-CAP.
[12] Jannink J.1999. Thesaurus entry extraction from an online dictionary [J]. In: Proceedings of Fusion‘99. Sunnyvale, Canada.
[13] Suryanto H, Compton P. 2001.Discovery of ontologies from knowledge bases [J]. In: Proceedings of the First International Conference on Knowledge Capture. British Columbia,Canada,171-178.
[14] Deitel A, Faron C, Dieng R.2001. Learning ontologies from RDF annotations [J]. In: Proceedings of the IJCAI Workshop in Ontology Learning. Seattle, USA.
[15] Volz R, Oberle D, Staab S, et al.2003. OntoLiFT rototype. http://wonderweb.semanticweb.org /deliverables/documents/D11.pdf.
[16] Stojanovic L, STojanovic N, Volz R. 2002. Migrating data intersive web sites into the semantic web[J]. In: Proceedings of the Severteenth ACM Symposium on Applide Computing. Madrid, Spain, 1100-1107.
[17] Shamsfard M, Barforoush A. 2004. A learning ontologies from natural language texts [J]. Intern- atinal Journal Human-Computer Studies, 60: 17-63.
[18] Wille R.Restructuring lattice theory: an approach based on hierarchies of concepts [J]. In: I. Riva1. eds.Ordered Sets,Reidel, Dordrecht, l982. 445-470.
[19] Godin R. Incremental concept formation algorithm based on Galois (concept) lattices [J]. Computational Intelligence,1995,11(2):246-267.
[20] Ho T B. An approach to concept format ion based on formal concept analysis [J]. IEICE Trans. Information and Systems, 1995, E78-D(5):553-559.
[21] Nourine L, Raynaud O. A fast algorithm for building lattices [J]. Work shop on Computational Graph Theory and Combinatorics [C]. Victoria, Canada, May, 1999.
[22] Carpineto C, Romano G. Galois: an order-theoretic approach to conceptual clustering [J]. Utgoff Ped. Proceedings of ICML-93 [C ].Amherst: Elsevier,1993. 33-40.
[23] Ho T B. Incremental conceptual clustering in the framework of Galois latt ice [J]. In Lu H, Motoda H, Liu H, eds. KDD: Techniques and Applications. World Scientific, 1997. 49-64.
[24] Baader F. Computing a minimal representation of the subsumption lattice of all conjunctions of concept defined in a terminology [J]. In: Ellis.G, Levinson R A, Fall, A, Dahl, V, eds. Proceedings of international KRUSE Symposium,UCSC,Santa Cruz,Aug,1995 168-178.
[25] Prediger S, Stumme G. Theory-driver logical scaling. In: Franconi, E, et al. eds [J]. Proceedings of the Sixth International Workshop Knowledge Representation Meets Databases. CEUR Workshop Proc. Vol.21, 1999.
[26] Mineau G, Stumme G, Wille R. Conceptual structures represented by conceptual graphs and formal concept analysis [J]. In: Tepfenbart, W, Cyre W, eds. Conceptual Structures: Stadards and Practices. Proc. Of ICCS 99. LNAI, Springer, Heidelberg, Vol.1640, 1999. 423-441.
[27] Obitko M, Snasel V, Smid J. Ontology Design with Formal Concept Analysis [J]. Edited by Vaclav snasel, Radim Belohlavek. In: Proc, of the CLA 2004 Intl. workshop on Concept Lattices and their Applications Ostrava, Czech Republic, Sep. 2004. 111-119.
[28] Haav H M. A Semi-automatic Mathod to Ontology Design by Using FCA [M]. Edited by Vaclav Snasel, Radim Belohlavek, In: Proc. Of the CLA 2004 Intl. Workshop on Concept Lattices and their Applications Ostrava , Czech Republic, Sep. 2004 13-24.
[29] Clark H H. Arenas of Language Use, chapter Common Ground and Language Use [J], Definite Reference and Mutual Knowledge. CSLI, 1992.
[30] Cimiano P, Stumme G,Hotho A, Tane J, Conceptual Knowledge Processing with Formal Concept Analysis and Ontologies [J]. In: Proc. Of the Second Intl. Conf. on formal concept analysis(ICFCA 04), Springer,2004.189-207.
[31]周文,刘宗田,陈慧田. FCA与本体结合研究的综述[J].计算机科学2006 Vol.33 No.2. 8-12.
[32] Gu Tao. Using Formal Concept Analysis for Ontology Structuring and Building [J]. ICIS, Nanyang Technological University, 2003.
[33] http://protégé.stanford.edu.
[34] http://sourceforge.net/projects/conexp.
[35]周文,刘宗田,陈慧琼.FCA与本体结合研究的综述[J].计算机科学2006 Vol.33No.2.8-12.
[36] Neto J L,Santo A D.Document clustering and text summarization [J]. In:Proc.of 4th Int1.Conf. on Practical Applications of Knowledge Discovery and Data Mining (PADI)_2000),2000.41-55.
[37] http://zh.wikipedia.org/wiki/
[38] Suk-Hyung Hwang, Hong-Gee Kim, and Hae-Sool Yang. A FCA-Based Ontology Construction for theDesign of Class Hierarchy [J]. O. Gervasi et al. (Eds.): ICCSA 2005, LNCS 3482, pp.2005: 827-835.
[39] R.Raina, A.Ng, and D. Koller. Constructing informative priors using transfer learning [J]. In ICML, 2006.
[40]王皓高阳,陈兴国.强化学习中的迁移:方法和进展[J].电子学报. 2008,第12A期,39-43.
[41]张文修,仇国芳.基于粗糙集的不确定决策[M].清华大学出版社.184-187.
[42] Alexander Maedche, Steffen Staab.Ontology Learning for the Semantic Web [J]. IEEE intelligent systems. 2001. 72-79.
[2]邓志鸿,唐世渭,张铭等.Ontology研究综述[J].北京大学学报:自然科学版, 2002(5):730-738.
[3]陈静,朱巧明,贡正仙.基于Ontology的信息抽取研究综述[J].计算机技术与发展, 2007(10):84-86.
[4] Nguyen N, Matsumoto M.Ontology-Based Context Management Agent for Vertical Handoff Use Fuzzy Logic Decision in Heterogeneous Network [C]. Proceedings of the First International Conference on Complex, Intelligent and Software Intensive Systems, 2006: 215-220.
[5] Li Jian, Jin Beihong. Query Division and Reformulation in Ontology-Based Heterogeneous Information Integration[C]. Proceedings of the15th International Conference on Computing, 2006:186-196.
[6]MaedcheA, Staab S. Ontology learning for the semantic web [J]. In: IEEE Intelligent Systems, special Issue on the Semantic Web.
[7] Gupta K M, Aha D W, Marsh E, et al. An architecture for engineering sublanguage WordNets [J]. In: Proceedings of the First International Conference on Global WordNet.
[8] Maedche A, Volz R. The text-to-onto ontology extraction and maintance environment [J]. In: Proceedings of the ICDM Workshop on Integrating Data Mining and Knowledge Management. California, USA.
[9] Khan L, Luo F. 2002.Ontology construction of information selection [J]. In: Proceedings of the Fourteenth IEEE International Conference on Tools with Artificial Intelligence. Washington D.C. USA.
[10] Kietz J, Maedche A, Volz R. 2000. A method for semi-automatic ontology acquisition from a corporate intranet[J]. In: Proceedings of EKAW'00 Workshop on Ontologies and Texts.
[11] Hahn U,Marko K G..2001. Joint knowledge capture for grammars and ontologies [J].In:Proceedings of the First International Conference on Knowledge Capture K-CAP.
[12] Jannink J.1999. Thesaurus entry extraction from an online dictionary [J]. In: Proceedings of Fusion‘99. Sunnyvale, Canada.
[13] Suryanto H, Compton P. 2001.Discovery of ontologies from knowledge bases [J]. In: Proceedings of the First International Conference on Knowledge Capture. British Columbia,Canada,171-178.
[14] Deitel A, Faron C, Dieng R.2001. Learning ontologies from RDF annotations [J]. In: Proceedings of the IJCAI Workshop in Ontology Learning. Seattle, USA.
[15] Volz R, Oberle D, Staab S, et al.2003. OntoLiFT rototype. http://wonderweb.semanticweb.org /deliverables/documents/D11.pdf.
[16] Stojanovic L, STojanovic N, Volz R. 2002. Migrating data intersive web sites into the semantic web[J]. In: Proceedings of the Severteenth ACM Symposium on Applide Computing. Madrid, Spain, 1100-1107.
[17] Shamsfard M, Barforoush A. 2004. A learning ontologies from natural language texts [J]. Intern- atinal Journal Human-Computer Studies, 60: 17-63.
[18] Wille R.Restructuring lattice theory: an approach based on hierarchies of concepts [J]. In: I. Riva1. eds.Ordered Sets,Reidel, Dordrecht, l982. 445-470.
[19] Godin R. Incremental concept formation algorithm based on Galois (concept) lattices [J]. Computational Intelligence,1995,11(2):246-267.
[20] Ho T B. An approach to concept format ion based on formal concept analysis [J]. IEICE Trans. Information and Systems, 1995, E78-D(5):553-559.
[21] Nourine L, Raynaud O. A fast algorithm for building lattices [J]. Work shop on Computational Graph Theory and Combinatorics [C]. Victoria, Canada, May, 1999.
[22] Carpineto C, Romano G. Galois: an order-theoretic approach to conceptual clustering [J]. Utgoff Ped. Proceedings of ICML-93 [C ].Amherst: Elsevier,1993. 33-40.
[23] Ho T B. Incremental conceptual clustering in the framework of Galois latt ice [J]. In Lu H, Motoda H, Liu H, eds. KDD: Techniques and Applications. World Scientific, 1997. 49-64.
[24] Baader F. Computing a minimal representation of the subsumption lattice of all conjunctions of concept defined in a terminology [J]. In: Ellis.G, Levinson R A, Fall, A, Dahl, V, eds. Proceedings of international KRUSE Symposium,UCSC,Santa Cruz,Aug,1995 168-178.
[25] Prediger S, Stumme G. Theory-driver logical scaling. In: Franconi, E, et al. eds [J]. Proceedings of the Sixth International Workshop Knowledge Representation Meets Databases. CEUR Workshop Proc. Vol.21, 1999.
[26] Mineau G, Stumme G, Wille R. Conceptual structures represented by conceptual graphs and formal concept analysis [J]. In: Tepfenbart, W, Cyre W, eds. Conceptual Structures: Stadards and Practices. Proc. Of ICCS 99. LNAI, Springer, Heidelberg, Vol.1640, 1999. 423-441.
[27] Obitko M, Snasel V, Smid J. Ontology Design with Formal Concept Analysis [J]. Edited by Vaclav snasel, Radim Belohlavek. In: Proc, of the CLA 2004 Intl. workshop on Concept Lattices and their Applications Ostrava, Czech Republic, Sep. 2004. 111-119.
[28] Haav H M. A Semi-automatic Mathod to Ontology Design by Using FCA [M]. Edited by Vaclav Snasel, Radim Belohlavek, In: Proc. Of the CLA 2004 Intl. Workshop on Concept Lattices and their Applications Ostrava , Czech Republic, Sep. 2004 13-24.
[29] Clark H H. Arenas of Language Use, chapter Common Ground and Language Use [J], Definite Reference and Mutual Knowledge. CSLI, 1992.
[30] Cimiano P, Stumme G,Hotho A, Tane J, Conceptual Knowledge Processing with Formal Concept Analysis and Ontologies [J]. In: Proc. Of the Second Intl. Conf. on formal concept analysis(ICFCA 04), Springer,2004.189-207.
[31]周文,刘宗田,陈慧田. FCA与本体结合研究的综述[J].计算机科学2006 Vol.33 No.2. 8-12.
[32] Gu Tao. Using Formal Concept Analysis for Ontology Structuring and Building [J]. ICIS, Nanyang Technological University, 2003.
[33] http://protégé.stanford.edu.
[34] http://sourceforge.net/projects/conexp.
[35]周文,刘宗田,陈慧琼.FCA与本体结合研究的综述[J].计算机科学2006 Vol.33No.2.8-12.
[36] Neto J L,Santo A D.Document clustering and text summarization [J]. In:Proc.of 4th Int1.Conf. on Practical Applications of Knowledge Discovery and Data Mining (PADI)_2000),2000.41-55.
[37] http://zh.wikipedia.org/wiki/
[38] Suk-Hyung Hwang, Hong-Gee Kim, and Hae-Sool Yang. A FCA-Based Ontology Construction for theDesign of Class Hierarchy [J]. O. Gervasi et al. (Eds.): ICCSA 2005, LNCS 3482, pp.2005: 827-835.
[39] R.Raina, A.Ng, and D. Koller. Constructing informative priors using transfer learning [J]. In ICML, 2006.
[40]王皓高阳,陈兴国.强化学习中的迁移:方法和进展[J].电子学报. 2008,第12A期,39-43.
[41]张文修,仇国芳.基于粗糙集的不确定决策[M].清华大学出版社.184-187.
[42] Alexander Maedche, Steffen Staab.Ontology Learning for the Semantic Web [J]. IEEE intelligent systems. 2001. 72-79.