Message Board

Dear readers, authors and reviewers,you can add a message on this page. We will reply to you as soon as possible!

2022 Volume 47 Issue 4
Article Contents

WEI Ling, WANG Zhen, QI Jianjun, et al. Necessity-Possibility Semi-Three-Way Concept[J]. Journal of Southwest China Normal University(Natural Science Edition), 2022, 47(4): 12-20. doi: 10.13718/j.cnki.xsxb.2022.04.002
Citation: WEI Ling, WANG Zhen, QI Jianjun, et al. Necessity-Possibility Semi-Three-Way Concept[J]. Journal of Southwest China Normal University(Natural Science Edition), 2022, 47(4): 12-20. doi: 10.13718/j.cnki.xsxb.2022.04.002

Necessity-Possibility Semi-Three-Way Concept

More Information
  • Received Date: 29/12/2021
    Available Online: 20/04/2022
  • MSC: O29;TP18

  • Three-way concept analysis, a combination of formal concept analysis and three-way decision, is a new tool for data analysis and knowledge discovery. Three-way concept has strict semantics and describes positive information and negative information from the same perspective, which limit its applications in practical problems, such as teamwork. Based on that, by combining necessity operator and possibility operator, necessity-possibility three-way operators are firstly defined by using interval set in this paper, and their properties are researched. Further, necessity-possibility semi-three-way concepts are proposed through introducing the thought of semiconcept into three-way concept analysis, and their algebraic structures are also studied. Necessity-possibility semi-three-way concepts can describe information from multi-perspective, which extends the semantics of three-way concept.
  • 加载中
  • [1] WILLE R. Restructuring Lattice Theory: An Approach Based on Hierarchies of Concepts[M]//RIVAL I. Ordered Sets. Dordrecht: Springer, 1982: 445-470.

    Google Scholar

    [2] GANTER B, WILLE R. Formal Concept Analysis: Mathematical Foundations[M]. New York: Springer-Verlag, 1999.

    Google Scholar

    [3] 张文修, 魏玲, 祁建军. 概念格的属性约简理论与方法[J]. 中国科学E辑: 信息科学, 2005, 35(6): 628-639.

    Google Scholar

    [4] 魏玲, 祁建军, 张文修. 决策形式背景的概念格属性约简[J]. 中国科学E辑: 信息科学, 2008, 38(2): 195-208. doi: 10.3321/j.issn:1006-9275.2008.02.003

    CrossRef Google Scholar

    [5] LI J H, MEI C L, LV Y J. Knowledge Reduction in Decision Formal Contexts[J]. Knowledge-Based Systems, 2011, 24(5): 709-715. doi: 10.1016/j.knosys.2011.02.011

    CrossRef Google Scholar

    [6] WU W Z, LEUNG Y, MI J S. Granular Computing and Knowledge Reduction in Formal Contexts[J]. IEEE Transactions on Knowledge and Data Engineering, 2009, 21(10): 1461-1474. doi: 10.1109/TKDE.2008.223

    CrossRef Google Scholar

    [7] 黄治国, 杨清琳. 基于启发式二分策略的属性约简方法[J]. 西南师范大学学报(自然科学版), 2020, 45(4): 59-67.

    Google Scholar

    [8] LI J H, MEI C L, KUMAR C A, et al. On Rule Acquisition in Decision Formal Contexts[J]. International Journal of Machine Learning and Cybernetics, 2013, 4(6): 721-731. doi: 10.1007/s13042-013-0150-z

    CrossRef Google Scholar

    [9] QI J J, WEI L, YAO Y Y. Three-Way Formal Concept Analysis[C]//MIAO D, PEDRYCZ W, S ' LęZAK D, et al. Rough Sets and Knowledge Technology RSKT 2014: Lecture Notes in Computer Science vol 8818. Switzerland: Springer, 2014: 732-741.

    Google Scholar

    [10] QI J J, QIAN T, WEI L. The Connections Between Three-Way and Classical Concept Lattices[J]. Knowledge-Based Systems, 2016, 91: 143-151. doi: 10.1016/j.knosys.2015.08.006

    CrossRef Google Scholar

    [11] 智慧来, 徐彤, 李逸楠. 基于三支概念簇的知识表示[J]. 西南大学学报(自然科学版), 2021, 43(10): 10-18.

    Google Scholar

    [12] WILLE R. The Basic Theorem of Triadic Concept Analysis[J]. Order, 1995, 12(2): 149-158. doi: 10.1007/BF01108624

    CrossRef Google Scholar

    [13] WAN Q, LI J H, WEI L. Optimal Granule Combination Selection Based on Multi-Granularity Triadic Concept Analysis[J]. Cognitive Computation, 2021, 2021: 1-15.

    Google Scholar

    [14] WEI L, QIAN T, WAN Q, et al. A Research Summary About Triadic Concept Analysis[J]. International Journal of Machine Learning and Cybernetics, 2018, 9(4): 699-712. doi: 10.1007/s13042-016-0599-7

    CrossRef Google Scholar

    [15] CHEN J H, ZHENG H Y, WEI L, et al. Factor Diagnosis and Future Governance of Dangerous Goods Accidents in China's Ports[J]. Environmental Pollution, 2020, 257: 1-8.

    Google Scholar

    [16] QUINTERO N Y, RESTREPO G. Formal Concept Analysis Applications in Chemistry: From Radionuclides and Molecular Structure to Toxicity and Diagnosis[M]. New-York: Springer, 2017.

    Google Scholar

    [17] XIE J P, YANG M H, LI J H, et al. Rule Acquisition and Optimal Scale Selection in Multi-Scale Formal Decision Contexts and Their Applications to Smart City[J]. Future Generation Computer Systems, 2018, 83: 564-581. doi: 10.1016/j.future.2017.03.011

    CrossRef Google Scholar

    [18] YAO Y Y. An Outline of A Theory of Three-Way Decisions[C]//YAO J T, YANG Y, SLOWINSKIN, et al. Proceedings of 2012 Rough Sets and Current Trends in Computing: Lecture Notes in Computer Science vol 7413. Berlin, Heidelberg: Springer, 2012: 1-17.

    Google Scholar

    [19] CIUCCI D. Orthopairs: A Simple and Widely Used Way to Model Uncertainty[J]. Fundamenta Informaticae, 2011, 108(3/4): 287-304.

    Google Scholar

    [20] REN R S, WEI L. The Attribute Reductions of Three-Way Concept Lattices[J]. Knowledge-Based Systems, 2016, 99: 92-102. doi: 10.1016/j.knosys.2016.01.045

    CrossRef Google Scholar

    [21] WANG Z, WEI L, QI J, et al. Attribute Reduction of SE-ISI Concept Lattices for Incomplete Contexts[J]. Soft Computing, 2020, 24(20): 15143-15158. doi: 10.1007/s00500-020-05271-2

    CrossRef Google Scholar

    [22] QIAN T, WEI L, QI J J. Constructing Three-Way Concept Lattices Based on Apposition and Subposition of Formal Contexts[J]. Knowledge-Based Systems, 2017, 116: 39-48. doi: 10.1016/j.knosys.2016.10.033

    CrossRef Google Scholar

    [23] YANG S C, LU Y N, JIA X Y, et al. Constructing Three-Way Concept Lattice Based on The Composite of Classical Lattices[J]. International Journal of Approximate Reasoning, 2020, 121: 174-186. doi: 10.1016/j.ijar.2020.03.007

    CrossRef Google Scholar

    [24] 祁建军, 汪文威. 多线程并行构建三支概念[J]. 西安交通大学学报, 2017, 51(3): 116-121.

    Google Scholar

    [25] WEI L. LIU L, QI J J, et al. Rules Acquisition of Formal Decision Contexts Based on Three-Way Concept Lattices[J]. Information Sciences, 2020, 516: 529-544. doi: 10.1016/j.ins.2019.12.024

    CrossRef Google Scholar

    [26] HUANG C C, LI J H, MEI C L, et al. Three-Way Concept Learning Based on Cognitive Operators: An Information Fusion Viewpoint[J]. International Journal of Approximate Reasoning, 2017, 83(1): 218-242.

    Google Scholar

    [27] YU H Y, LI Q G, CAI M J. Characteristics of Three-Way Concept Lattices and Three-Way Rough Concept Lattices[J]. Knowledge-Based Systems, 2018, 146: 181-189. doi: 10.1016/j.knosys.2018.02.007

    CrossRef Google Scholar

    [28] ZHAO X R, MIAO D Q, HU B Q. On Relationship Between Three-Way Concept Lattices[J]. Information Sciences, 2020, 538: 396-414. doi: 10.1016/j.ins.2020.06.007

    CrossRef Google Scholar

    [29] VORMBROCK B, WILLE R. Semiconcept and Protoconcept Algebras: The Basic Theorems[M]//Formal Concept Analysis. Berlin, Heidelberg: Springer Berlin Heidelberg, 2005: 34-48.

    Google Scholar

    [30] HOWLADER P, BANERJEE M. Algebras from Semiconcepts in Rough Set Theory[C]//NGUYEN H, HA Q T, LI T, et al. International Joint Conference on Rough Sets. Switzerland: Springer Cham, 2018: 440-454.

    Google Scholar

    [31] YAO Y Y. Interval Sets and Interval-Set Algebras[C]//2009 8th IEEE International Conference on Cognitive Informatics. June 15-17, 2009, Hongkong China. Piscataway: IEEE, 307-314.

    Google Scholar

    [32] YAO Y Y. Three-Way Decision and Granular Computing[J]. International Journal of Approximate Reasoning, 2018, 103: 107-123. doi: 10.1016/j.ijar.2018.09.005

    CrossRef Google Scholar

    [33] DVNTSCH N, GEDIGA G. Modal-Style Operators in Qualitative Data Analysis[C]//2002 IEEE International Conference on Data Mining, December 9-12, 2002, Maebashi City, Japan, Piscataway: IEEE, 2002: 155-162.

    Google Scholar

    [34] DVNTSCH I, GEDIGA G. Approximation Operators in Qualitative Data Analysis[C]//SWART H, ORłOWSKA E, SCHMIDT G, et al. Theory and Applications of Relational Structures as Knowledge Instruments. Lecture Notes in Computer Science vol 2929. Berlin Heidelberg: Springer, 2003: 214-230.

    Google Scholar

    [35] 李进金, 孙文. 知识空间、形式背景和知识基[J]. 西北大学学报(自然科学版), 2019, 49(4): 517-526.

    Google Scholar

    [36] SUN W, LI J J, GE X, et al. Knowledge Structures Delineated by Fuzzy Skill Maps[J]. Fuzzy Sets and Systems, 2021, 407: 50-66. doi: 10.1016/j.fss.2020.10.004

    CrossRef Google Scholar

    [37] PEI D W, XU Z B. Rough Set Models on Two Universes[J]. International Journal of General Systems, 2004, 33(5): 569-581. doi: 10.1080/0308107042000193561

    CrossRef Google Scholar

    [38] 谢德华, 刘财辉, 凌敏. 局部多粒度覆盖粗糙集[J]. 西南大学学报(自然科学版), 2021, 43(10): 1-9.

    Google Scholar

    [39] SHAO M W. GUO L, WANG C Z. Connections Between Two-Universe Rough Sets and Formal Concepts[J]. International Journal of Machine Learning and Cybernetics, 2018, 9(11): 1869-1877. doi: 10.1007/s13042-018-0803-z

    CrossRef Google Scholar

    [40] ZHI H L, QI J J. Common-Possible Concept Analysis: A Granule Description Viewpoint[J]. Applied Intelligence, 2021, 2021: 1-12.

    Google Scholar

  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(4)  /  Tables(3)

Article Metrics

Article views(4217) PDF downloads(211) Cited by(0)

Access History

Other Articles By Authors

Necessity-Possibility Semi-Three-Way Concept

Abstract: Three-way concept analysis, a combination of formal concept analysis and three-way decision, is a new tool for data analysis and knowledge discovery. Three-way concept has strict semantics and describes positive information and negative information from the same perspective, which limit its applications in practical problems, such as teamwork. Based on that, by combining necessity operator and possibility operator, necessity-possibility three-way operators are firstly defined by using interval set in this paper, and their properties are researched. Further, necessity-possibility semi-three-way concepts are proposed through introducing the thought of semiconcept into three-way concept analysis, and their algebraic structures are also studied. Necessity-possibility semi-three-way concepts can describe information from multi-perspective, which extends the semantics of three-way concept.

  • 形式概念分析(Formal concept analysis,FCA)作为有效的、极具潜力的知识发现工具,于1982年由德国数学家Wille首次提出,用于概念的发现、排序和显示[1-2]. Wille将数据描述为形式背景,并定义了一对导出算子,由此生成形为(外延,内涵)二元对的形式概念,继而生成作为该理论核心数据结构的概念格. 概念格是根据形式背景中对象与属性之间的二元关系建立的一种概念层次结构,生动简洁地体现了概念之间的泛化与特化关系. 目前FCA的主要研究方向有属性约简[3-7]、规则获取[5, 8]以及向三支概念分析[9-11]、三元概念分析[12-14]的拓广等. 同时该理论还在港口危险品管理、智慧城市等领域得到了成功的应用[15-17].

    三支概念分析(Three-way concept analysis,3WCA)是将三支决策理论(Three-way decision,3WD)[18]中的三分思想应用于FCA而新产生的一种进行知识发现的理论[9-10]. 与形式概念一样,三支概念也表现为(外延,内涵)二元对的形式,但区别于形式概念,三支概念的内涵/外延为一个正交对[19],用于刻画属性集/对象集的三分. 因此3WCA既具有FCA的基本表现形式和具体研究内容,也体现了3WD的“三分而治”思想. 自2014年提出以来,其思想逐渐为知识发现研究领域的研究人员所接受,研究主题与理论成果也越来越多,如:三支概念格属性约简[20-21]、三支概念格构建[22-24]、三支规则获取[25]、三支概念学习[26]、三支概念格的特征分析[27]以及三支概念格之间的关系研究[28]等.

    三支概念从共性这一角度,同时考虑对象与属性之间“共同具有”以及“共同不具有”正、负两方面的信息. 但在某些实际问题中,共性信息可能不那么重要,比如在团队合作问题中,我们不仅需要知道一个团队可以合作完成哪些任务,还需要关注哪些任务仅可由该团队的成员完成,这就需要其他模态算子为我们提供的诸如必然信息等其他视角的信息,因此结合不同的模态算子来实现信息的多视角刻画就十分有意义. 其次,三支概念的外延与内涵存在着较为严苛的双向对应,故三支概念的获取也较为复杂,因此将FCA中半概念[29-30]的单向对应思想引入3WCA也显得十分有意义. 基于此,本文结合必然算子与可能算子,利用区间集[31]提出必然-可能三支算子,并定义必然-可能半三支概念,进而生成必然-可能半三支概念格.

1.   基础知识
  • 本节给出正交对、区间集以及3WCA的一些基本概念与性质.

    对于任意非空论域UU的幂集记为$\mathscr{D}$(U),U的幂集的笛卡尔积记为$\mathscr{D}\mathscr{P}$(U),即

    对于任意的(A1B1),(A2B2)∈$\mathscr{D}\mathscr{P}$(U),$\mathscr{D}\mathscr{P}$(U)上的包含、并、交分别定义为

    特别地,对于(AB)∈$\mathscr{D}\mathscr{P}$(U),如果AB=Ø,那么称(AB)为一个正交对.

    U上的区间集为

    U上所有的区间集记为$\mathscr{I}\mathscr{P}$(U). 对于任意[A1A1],[A2A2]∈ $\mathscr{I}\mathscr{P}$ (U),$\mathscr{I}\mathscr{P}$(U)上的包含、并、交分别定义为

    3WCA的数据基础是形式背景.

    定义1[2]  称三元组(GMI)为一个形式背景,其中G={g1g2,…,gp}为对象集,每个gi(ip)称为一个对象; M={m1m2,…,mq}为属性集,每个mj(jq)称为一个属性. IG×MGM之间的二元关系,若(gm)∈I,则称对象g具有属性m.

    在形式背景(GMI)下,定义*算子与*算子为:对于任意的对象子集XG和属性子集AM

    其中c表示集合的补集,即Ic=G×MI.

    特别地,对于任意的gGmM,记{g}*g*,{m}*m*.

    结合*算子与*算子,3WCA定义一对对象导出三支算子,并可形成对象导出三支概念.

    定义2[9-10]  设(GMI)是一个形式背景. 对于任意的对象子集XG和属性子集ABM,一对对象导出三支算子,即OE-算子$^\lessdot$$\mathscr{P} $(G) DP (M),$^\gtrdot$$\mathscr{D}\mathscr{P}$(M) $\mathscr{P} $(G)定义如下:

    X$^\lessdot$=(AB)且(AB)$^\gtrdot$=X同时成立,则(X,(AB))称为(GMI)的一个对象导出三支概念,简称为OE-概念,其中X称为OE-概念的外延,(AB)称为OE-概念的内涵.

    对于对象子集Ø≠XG,可以利用OE-算子得到属性集M上的一个正交对(X*X*),这个正交对可将M分为3部分:正域POSX=X*、负域NEGX=X*,以及中间域BNDX=M-(X*X*). POSX,NEGX和BNDX两两互不相交,形成M的一个含空集的三划分,即弱三划分[32].

    类似地,对于任意的属性子集AM和对象子集XYG,一对属性导出三支算子,简称AE-算子,$^\lessdot$$\mathscr{P} $(M) $\mathscr{D}\mathscr{P}$(G),$^\gtrdot$$\mathscr{D}\mathscr{P}$(G) $\mathscr{P} $(M),定义如下:

    ((XY),A))称为(GMI)的一个属性导出三支概念,即AE-概念,当且仅当(XY)$^\gtrdot$=AA$^\lessdot$=(XY) 同时成立. (XY)称为AE-概念的外延,A称为AE-概念的内涵.

    所有OE-概念组成的集合OEL(GMI)可形成一个完备格,称之为对象导出三支概念格(简称OE-概念格). 对偶地,所有AE-概念形成的完备格称为属性导出三支概念格AEL(GMI)(简称AE-概念格). OE-概念格及AE-概念格上的偏序关系及上、下确界具体可参见文献[9-10].

    例1  表 1为形式背景(GMI),其中对象集G={1,2,3,4}为4名学生的集合,属性集M= {abcd}为4个问题的集合. 表 1中“1”表示学生能解决该问题,“0”则表示不能,如:学生1能解决问题acd,不能解决问题b. 表 1形式背景的OE-概念格、AE-概念格分别如图 1图 2所示.

    下面以(13,(cb))为例解释OE-概念的语义. 该概念表明:学生1与3都可以解决的问题是c,都不能解决的问题是b,而能解决c且不能解决问题b的学生也恰为1与3. {c}∪{b} 的补集中的问题ad则是学生1与3具有差异的问题. 所以OE-概念清晰、完整、准确地反映了一个对象子集的正、负两种共性与差异性.

    AE-概念的解释与OE-概念类似,只不过是从属性角度出发,考虑一个属性子集被对象子集共同具有以及共同不具有的情况.

    此外,一些研究人员从模态逻辑的角度,将*算子称为充分算子,并陆续提出必然算子、可能算子以及对偶充分算子等概念[29-30].

    给定形式背景(GMI),对于任意XGAM,必然算子定义为

    可能算子定义为

    以例1中的对象子集{1,2}为例,{1,2}={d}表明:能解决问题d的学生一定在学生1与2之中,{1,2}=M表明:学生1和2能解决所有问题.

    必然算子与可能算子的性质如下:

    性质1[33-34]  设(GMI)为一个形式背景,对于任意XX1X2GAA1A2M,有:

    (i) XXAA;

    (ii) X1X2X1X2A1A2A1A2;

    (iii) X1X2X1X2A1A2A1A2;

    (iv) (X1X2)X1X2,(A1A2)A1A2;

    (v) (X1X2)=X1X2,(A1A2)=A1A2;

    (vi) (X1X2)=X1X2,(A1A2)=A1A2;

    (vii) (X1X2)X1X2,(A1A2)A1A2.

2.   对象导出必然可能半三支概念
  • 3WCA从正、负两方面同时考虑共性,利用正交对实现对象集\属性集的三分. 事实上,不同的模态算子从不同的角度刻画不同的信息,并且不同模态算子之间也存在着内在关系,如:必然算子与可能算子间的包含关系,使得我们可以结合两种不同的模态算子,利用区间集等工具,定义新的三支算子来获取多视角信息,并实现对象集\属性集的三分.

  • 结合必然算子与可能算子,从对象子集三分属性集的角度,定义对象导出必然-可能三支算子如下:

    定义3  设(GMI)为一个形式背景,对于任意X$\mathscr{P} $(G),对象导出必然-可能三支算子$^\vartriangleleft$$\mathscr{P} $(G) $\mathscr{I}\mathscr{P}$(M),定义为X$^\vartriangleleft$=[XX],简称为ONPE-算子.

    ONPE-算子具有如下性质:

    性质2  设(GMI)为一个形式背景,对于任意X1X2$\mathscr{P} $(G),我们有:

    (i) X1X2X1$^\vartriangleleft$X2$^\vartriangleleft$;

    (ii) (X1X2)$^\vartriangleleft$X1$^\vartriangleleft$X2$^\vartriangleleft$;

    (iii)(X1X2)$^\vartriangleleft$X1$^\vartriangleleft$X2$^\vartriangleleft$.

      (i) 若X1X2,由性质1(b),(c),我们有X1X2X1X2,则

    (ii) 由性质1(d),(e),可知

    (iii) 由性质1(f),(g),我们有

  • 形式概念分析中,称形如(XX*),(A*A)的二元对为形式背景(GMI)的半概念,其中XG是任意的对象子集,AM是任意的属性子集. 不难看出,半概念仅考虑对象子集与属性子集间的单向对应,而不再强调苛刻的双向对应. 基于这种思想,我们利用ONPE-算子,定义对象导出必然-可能半三支概念为:

    定义4  设(GMI)为一个形式背景,对于任意X$\mathscr{P} $(G),[AA]∈$\mathscr{I}\mathscr{P}$(M),若X$^\vartriangleleft$=[AA],则称(X,[AA])为对象导出必然-可能半三支概念,简称为ONPSE-概念,其中X称为ONPSE-概念的外延,[AA]称为ONPSE-概念的内涵.

    对于对象子集XG,区别于OE-算子获取的共有属性,ONPE-算子可以同时获取必然属性X与可能属性X,并形成属性集M上的一个区间集[XX],这个区间集可将M分为3部分:正域POSX=X、负域NEGX=M-X以及中间域BNDX=XX. 并且POSX,NEGX和BNDX两两互不相交,形成M的一个弱三划分.

    例2  表 1形式背景下所有的ONPSE-概念如表 2所示.

    以(12,[dM]),(14,[dM])为例解释ONPSE-概念的语义. 概念(12,[dM])表明:能解决问题d的学生在学生1与学生2当中,学生1和学生2可以合作解决所有的问题. 类似地,概念(14,[dM])表明:能解决问题d的学生在学生1与学生4当中,学生1和学生4可以合作解决所有的问题. 结合两个ONPSE-概念,我们进一步可知,能解决问题d的学生是学生1,符合概念(1,[dacd])所反映的信息.

    记(GMI)的所有ONPSE-概念的集合为ONPSEL(GMI),定义其偏序关系为:对于任意(X1,[A1A1]),(X2,[A2A2])∈ONPSEL(GMI),

    称ONPSEL(GMI)为(GMI)的对象导出必然-可能半三支概念格,简称为ONPSE-概念格. 定理1给出其上、下确界,并证明其是一个完备格.

    定理1 ONPSEL(GMI) 是一个完备格,其上、下确界分别为:

      对于任意的(X1,[A1A1]),(X2,[A2A2])∈ONPSEL(GMI),由ONPE-算子的定义可知

    故(X1X2,[(X1X2)A1A2])∈ONPSEL(UAI). 再证(X1X2,[(X1X2)A1A2]) 是上确界. 首先由X1X1X2X2X1X2,有

    则其为上界. 下证是最小上界. 设(X,[AA])是(X1,[A1A1]),(X2,[A2A2])的任意上界,即

    X1XX2X,故X1X2X,因此

    是最小上界,即上确界. 同理可证,(X1X2,[A1A2,(X1X2)])是一个ONPSE-概念,且为下确界.

    例3  表 1的ONPSE-概念格如图 3所示.

    由ONPSE-概念定义可知,任给一个对象子集XG,都存在一个ONPSE-概念(X,[AA])与之一一对应,因此有如下结论成立:

    定理2  ONPSEL(GMI)$ \cong $$\mathscr{P} $(G),其中$\mathscr{P} $(G)为G的幂集格.

3.   属性导出必然-可能半三支概念
  • 结合必然算子与可能算子,从属性子集三分对象集的角度研究属性导出必然-可能半三支概念. 因与第2节类似,本节证明省略. 首先,定义属性导出必然-可能三支算子如下:

    定义5  设(GMI)为一个形式背景,对于任意A$\mathscr{P} $(M),定义属性导出必然-可能三支算子$\vartriangleright$$\mathscr{P} $(M)→$\mathscr{D}\mathscr{P}$(G)为A$\vartriangleright$=[AA]. 简称为ANPE-算子.

    ANPE-算子具有如下性质:

    性质3  设(GMI)为一个形式背景,对于任意A1A2$\mathscr{P} $(M),我们有:

    (i) A1A2A1$\vartriangleright$A2$\vartriangleright$;

    (ii) (A1A2)$\vartriangleright$A1$\vartriangleright$A2$\vartriangleright$;

    (iii) (A1A2)$\vartriangleright$A1$\vartriangleright$A2$\vartriangleright$.

  • ANPE-算子可以生成属性导出必然-可能半三支概念,简称ANPSE-概念.

    定义6  设(GMI)为一个形式背景,对于任意[XX]∈$\mathscr{I}\mathscr{P}$(G),A$\mathscr{P} $(M),若[XX]=A$\vartriangleright$,则称([XX],A)为属性导出必然-可能半三支概念,其中[XX]称为ANPSE-概念的外延,A称为ANPSE-概念的内涵.

    ONPSE-概念与ANPSE-概念统称为必然-可能半三支概念,简称为NPSE-概念.

    类似于ONPE-算子,对于属性子集AM,可以利用ANPE-算子得对象集G上的一个区间集[AA],这个区间集可将G分为3部分:正域POSA=A、负域NEGA=GA以及中间域BNDA=AA. 并且POSA,NEGA和BNDA互不相交,形成G的一个弱三划分.

    记(GMI)的所有ANPSE-概念的集合为ANPEL(GMI),对于任意的([X1X1],A1),([X2X2],A2)∈ANPSEL(GMI),定义其偏序关系:

    称ANPSEL(GMI)为(GMI)的属性导出必然-可能半三支概念格,简称为ANPSE-概念格.

    定理3  ANPSEL(GMI)是一个完备格,其上、下确界分别为

    例4  表 1的所有ANPSE-概念如表 3所示,ANPSE-概念格如图 4所示.

    以([234,G],abc)为例解释ANPSE-概念的语义. 概念([234,G],abc)表明:问题abc可被所有学生合作解决,学生2,3,4能解决的问题在abc当中. 并且G与{2,3,4}的差集中的学生1还能解决其他问题,如概念([Ø,1],d)表明:学生1还能解决问题d.

    由ANPSE-概念定义,可得以下结论:

    定理4  ANPSEL(GMI)$ \cong $$\mathscr{P} $(M),其中$\mathscr{P} $(M)为M的幂集格.

4.   结语
  • 本文基于多视角信息获取这一想法,结合必然算子与可能算子,从对象或属性这两种不同的角度出发,定义了对象导出与属性导出必然-可能三支算子,并结合形式概念分析中半概念的单向对应思想,获取了对象导出与属性导出必然-可能半三支概念,拓广了三支概念的语义.

    事实上,对象与属性作为两种不同的研究角度,生成的两种必然-可能半三支概念的语义解释与应用场景也有所不同,如在形式概念分析与知识空间理论结合的研究[35-36]中,从对象或属性出发可分别看作是问题驱动或技能驱动这两种截然不同的研究方向,因此如何结合上述研究成果,对两种必然-可能半三支概念的语义与应用做进一步研究就很有意义. 本文仅从正面结合必然算子与可能算子获取了多视角的信息,而负信息也很重要,因此如何从负面结合必然算子与可能算子来获取信息也十分有意义. 并且必然算子与可能算子与双论域粗糙集[37-38]的下、上近似有着紧密联系[39],因此本文获取的必然-可能半三支概念对双论域粗糙集的知识可视化也有所帮助,如:我们可以从ONPSE-概念格迅速获取双论域粗糙集的可定义集与不可定义集,因此如何将本文研究内容与双论域粗糙集结合也很重要. 最后,其他模态算子的结合研究也十分有意义,如文献[40]结合充分算子与可能算子探讨了共同-可能粒描述. 这些都是我们未来的研究方向.

Figure (4)  Table (3) Reference (40)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return