Semantic Information Retrieval Based on Fuzzy Ontology for Electronic Commerce

20 JOURNAL OF SOFTWARE, VOL. 3, NO. 9, DECEMBER 2008

Jun Zhai, Yiduo Liang, Yi Yu and Jiatao Jiang School of Management, Dalian Maritime University, Dalian 116026, P. R. China [email protected]

Abstract—Information retrieval is the important work for forward as one of the motivations of the Semantic Web Electronic Commerce. Ontology-based semantic retrieval is [3]-[4]. a hotspot of current research. In order to achieve fuzzy However, the conceptual formalism supported by semantic retrieval, this paper applies a fuzzy ontology typical ontology may not be sufficient to represent framework to information retrieval system in E-Commerce. uncertainty information commonly found in many The framework includes three parts: concepts, properties of concepts and values of properties, in which property’s value application domains due to the lack of clear-cut can be either standard data types or linguistic values of boundaries between concepts of the domains. Moreover, fuzzy concepts. The semantic query expansions are fuzzy knowledge plays an important role in many constructed by order relation, equivalence relation, domains that face a huge amount of imprecise and vague inclusion relation, reversion relation and complement knowledge and information, such as text mining, relation between fuzzy concepts defined in linguistic multimedia information system, medical informatics, variable ontologies with Resource Description Framework machine learning, and human natural language processing (RDF). The application to retrieve customer, product and [5]. supplier information shows that the framework can To handle uncertainty of information and knowledge, overcome the localization of other fuzzy ontology models, and this research facilitates the semantic retrieval of one possible solution is to incorporate fuzzy theory into information through fuzzy concepts on the Semantic Web. ontology. Then we can generate fuzzy ontologies, which contain fuzzy concepts and fuzzy memberships. The Index Terms—semantic information retrieval, fuzzy fuzzy ontologies are capable of dealing with fuzzy ontology, ontology, electronic commerce, the Semantic Web knowledge [6], and are efficient in text and multimedia object representation and retrieval [7]. Lau [8] presented a fuzzy domain ontology for business knowledge I. INTRODUCTION management. Lee et al. [9] proposed an algorithm to create fuzzy ontology and applied it to news Along with the Internet fast development, the electronic commerce based on the network displays the summarization. Tho et al. proposed a Fuzzy Ontology more huge than traditional business advantage, raises the Generation Framework (FOGA) for fuzzy ontology performance and the efficiency of the traditional business generation on uncertainty information [10]. This framework is based on the idea of fuzzy theory and activity biggest. Existing electronic commerce system will be hard to solve growing the business information Formal Concept Analysis (FCA). Abulaish et al. [11]-[12] proposed a fuzzy ontology framework in which a concept explosion and the customer's characteristic need, so the customer who faces the amount of goods information descriptor is represented as a fuzzy relation which encodes the degree of a property value using a fuzzy hard to do a best choice, and lack to hand over with each other between the customer and the vender. membership function. Calegari and Ciucci [13] presented the fuzzy OWL language. Therefore, information retrieval (IR) is very important to achieve E-Commerce in WWW. Current information But, current fuzzy ontology models have localization in expressing uncertainty derived from ordinary fuzzy set, retrieval on the web is based primarily on keywords which often cause problems in precision and recall. and do not focus on essential semantic relationships between fuzzy concepts, which lead difficulty to search Ontology-based semantic retrieval is a hotspot of current research. Ontology is a conceptualization of a domain information at fuzzy semantic level. In order to achieve fuzzy semantic retrieval for E-Commerce, this paper into a human understandable, machine-readable format consisting of entities, attributes, relationships, and axioms applies a new kind of fuzzy ontology to information retrieval system in E-Commerce. The rest of this paper is [1]. It is used as a standard knowledge representation for the Semantic Web [2]. The use of ontologies to overcome organized as follows: Section 2 gives some basic definitions of intuitionistic fuzzy set. Section 3 introduces the limitations of keyword-based search has been put fuzzy domain ontology model. Section 4 proposes fuzzy linguistic variable ontology models and formal Corresponding author is Jun Zhai. representation with RDF. Section 5 presents fuzzy

ontology framework and section 6 applies the fuzzy (2) P is a set of concept properties. A property ontology to information retrieval for E-Commerce. p P is defined as an instance of a ternary relation of Finally, section 7 concludes the paper. the form p(c,v, f ) ,wherec C is an ontology II. BASIC DEFINITIONS OF INTUITIONISTIC FUZZY SET concept, v is a property value associated with c and f In this section, we review some fundamental defines restriction facets on v . Some of the restriction knowledge of fuzzy theory [14]. facets are type ( ft ), cardinality ( f c ), and range ( f r ). Definition 1 (Intuitionistic fuzzy set) – An The type facet f may be any one from the standard data intuitionistic fuzzy set A on a universe U is defined as t an object of the following form: types supported by ontology editors, i.e. ft {boolean, A {(u,P A (u),Q A (u)) | u U}, where the functions integer, float, string, symbol, instance, class, …}. The

P A (u) :U o[0,1] and Q A (u) :U o [0,1] define cardinality facet f c defines the upper and lower limits thedegreeofmembershipandthedegreeofnon- on the number of values for the property. The range facet membership of the element u U in A , respectively, fr specifies a range of values that can be assigned to the and for every u U : P A (u) Q A (u) d1 . property. The intuitionistic fuzzy set has an equivalent form of (3) R {r | r C uC} is a set of binary semantic interval value: A {(u,[P A(u),1Q A (u)]) | u U} , relations defined between concepts in C . Basic relations are defined as {synonym of, kind of, part of, instance of, where [P A (u),1Q A (u)] [0,1] . Obviously, property of} R . when , the intuitionistic fuzzy set is P A (u) Q A (u) 1 (4) A is a set of axioms. An axiom is a real fact or an ordinary fuzzy set. reasoning rule. Definition 2 (Intuitionistic fuzzy relation) – An Fuzzy ontology is created as an extension to the intuitionistic fuzzy relation from U to V is an standard ontology. intuitionistic fuzzy set on U uV : Definition 4 (Fuzzy domain ontology) – A fuzzy domain ontology is a 6-tuple R { (u,v),P R (u,v),Q R (u, v) !| (u,v)U uV }, C R O (I, C, P , R, P , A ) ,where: where P (u, v) : U uV o [0,1] , F F R (1) I is the set of individuals, also called instances of Q R (u,v) : U uV o [0,1], P R (u, v) Q R (u,v) d1. the concepts. The intuitionistic fuzzy relation has also an equivalent (2) C is a set of concepts. Every concept here has form of interval value: some properties whose value is fuzzy concept or fuzzy set. R { (u,v),[a,b] !| (u,v)U uV } ,whereAnd, every concept can have the degree of membership

[a,b] [0,1]. P C (i) : I o[0,1] and the degree of non- Because an intuitionistic fuzzy set provides more membership vC (i) : I o[0,1] of the i I in C . choices for the attribute description of an object and has C stronger ability to express uncertainty than an ordinary (3) P is a set of concepts properties. A property fuzzy set, it has gained extensive attention from the p C PC is defined as a 5-tuple of the academic circles and the circles of engineering and C technology. Presently, some science branches based on form p (c, vF , q F , f ,U) ,wherec C is an intuitionistic fuzzy set have appeared, such as ontology concept, vF represents property values , q F intuitionistic fuzzy set topology, intuitionistic fuzzy set models linguistic qualifiers, which can control or alter the logic etc. The intuitionistic fuzzy set has applied to lots of fields such as artificial intelligence, decision-making strength of a property value vF , f is the restriction analysis, pattern recognition, handling intelligence facets on vF ,andU is the universe of discourse. Both information and so on. vF and q F are the fuzzy concepts on U ,butq F

III. FUZZY DOMAIN ONTOLOGY MODEL changes the fuzzy degree of vF . For example, “price” is Gruber defines ontology as an explicit specification of a property of concept “service”. The value of “price” a conceptualization, i.e. an abstract and simplified may be either fuzzy concept “cheap” or fuzzy number representation of real-world entities [15]. An ontology “around 50”, and the linguistic qualifiers may be “very”, organizes domain knowledge in terms of concepts, “little”, “close to” etc. Therefore, the final value of properties, relations and axioms. “price” may be “very cheap” or “little expensive”. At the Definition 3 (Ontology) – An ontology is a 4-tuple same time, the property p C P C has also the non-fuzzy O (C, P, R, A) ,where: C form p (c,v, f ). (1)C is a set of concepts defined for the domain. A concept is often considered as a class in an ontology.

(4) R is a set of inter-concept relations between (3) R {r | r C F u CF } is a set of binary concepts. The relation type is not only the ordinary binary relations between concepts in C . A kind of relation is relation of r C uC , but also is the fuzzy relation and F the intuitionistic fuzzy relation from C to C . set relation RS {inclusion ( i.e. ), intersection, R disjointness, complement ( i.e. )}, and the other (5) P is a set of relations properties. Like concept ŉ relations are the order relation and equivalence properties, p R PR is defined as a 4-tuple of the form relation RO {d, t, } . C F and an order relation R p (c1 ,c2 ,r,s F ) ,wherec1, c2 C are ontology r compose the ordered structure CF ,r ! .Thereare concepts, r represents relation, and sF [0,1] or other semantic relations between concepts, such as semantic distance relation, semantic proximity relation s F [0,1]models relation strengths and has meaning of and semantic association relation etc. fuzzy set or intuitionistic fuzzy set on C uC ,whichcan (4) F is the set of membership functions on U , represent the strength of association between concept- which is isomorphic to CF . The corresponding element pairs c1,c2 ! . For instance, there is a relation of “loyalty” between “customer” and “brand”. The strength of F is M in definition 5, but F has also certain of “loyalty” can be 0.7, a fuzzy value, and can be structure or relations. [0.6,0.8], a interval value, i.e. intuitionistic fuzzy value, (5) S {s | s : C F u CF o C F } is a set of binary which express more abundant information about operators at C . These binary operators form the uncertainty. F mechanism of generating new fuzzy concepts. Basic (6) AF is a set of fuzzy rules. In a fuzzy system the set operators are the “union”, “intersection” and of fuzzy rules is used as knowledge base. “complement” etc., i.e. S {,,,} . C and The fuzzy domain ontology is used to model domain F expert knowledge. But, due to the lack of relationships S compose the algebra structure C F , S ! . between fuzzy concepts that can be the value of (6) U is the universe of discourse. properties, it is difficult to search information at semantic Modeling the linguistic qualifiers, we extend the fuzzy level. Consequently, we propose the fuzzy linguistic linguistic variable ontology as follows. variables ontology models. Definition 7 (Extended fuzzy ontology) – An extended fuzzyontologyisa9-tuple IV. FUZZY LINGUISTIC VARIABLE ONTOLOGY OF (ca ,C F , R,F , S,Q, O, L,U ) ,where: The fuzzy linguistic variables proposed by Zadeh are (1) have same interpretations as the basic of fuzzy knowledge and fuzzy system. ca ,C F , R, F, S,U Definition 5 (Fuzzy linguistic variable) – A fuzzy defined in definition 6. linguistic variable is a 4-tuple (X ,T, M,U) ,where: (2) Q is the set of the linguistic qualifiers, e.g. (1) X is the name of fuzzy linguistic variable, e.g. Q ={very, little, close to, …}. An qualifier q Q and a “price” or “speed” etc. fuzzy concept cF C F compose a composition fuzzy (2) T is the set of terms which is the value of fuzzy concept that can be the value of , e.g. “very cheap”. linguistic variable, e.g. T ={ cheap, appropriate, ca expensive, …} or T ={fast, middle, slow,…}. (3) O is the set of fuzzy operators on U ,whichis (3) M is the mapping rules which map every term of isomorphic to Q . T to fuzzy set on U . (4) L (Q uC F ) (C F uQ) is a binary relation (4) U is the universe of discourse. from Q to C or C to Q . Introducing semantic relationships between concepts, F F we obtain the ontology model. q, cF ! or cF , q !L mean that q Q and Definition 6 (Fuzzy linguistic variable ontology) – A c C can compose a composition fuzzy concept. fuzzy linguistic variable ontology is a 6-tuple F F To simplify the transform from fuzzy linguistic OF (ca ,C F ,R, F, S,U) ,where: variables to fuzzy ontology, we introduce the basic fuzzy ontology model as follows. (1) ca is a concept on the abstract level, e.g. “price”, Definition 8 (Basic fuzzy ontology) – A basic fuzzy “speed” etc. The corresponding element of ca is X in ontology is a 4-tuple OF (ca ,C F , F,U ) ,where definition 5. ca , CF , F,U have same interpretations as defined in (2) C F is the set of fuzzy concepts which describes all definition 6, which satisfy the following conditions: values of ca . The corresponding element of CF is T in (1) C F {c1 ,c2 ,,cn } is a limited set. definition 5, but C F has certain structure or relations R .

(2) Only one relation of set, the relation of disjointness, exists in C ,andC is complete on U . In the other F F (3) C F has an ordered relation d ,and C F ,d! is a complete ordered set, i.e. all concepts in C F constitute a An example of basic fuzzy ontology is OF ˙ Each fuzzy concept is associated with a membership ˄ ca price of product, C F {cheap, appropriate, function. There are many types of membership functions. expensive}, U [0,100] ˅ , where “cheap” Some of the common ones are: “appropriate” “expensive” , and the membership (1) Triangular. A triangular shaped curve can be functions are shown in Fig 1. described by three points, namely: (x1, 0), (x2, 1), and The Semantic Web, introduced by Tim Bemers-Lee, (x3, 0). The RDF statements are as following: uses Resource Description Framework (RDF) to add structure and meaning to Web applications. RDF data model “resource-property-value” is the current standards for establishing semantic interoperability on the Web [16]. Fig. 2 describes “basic fuzzy ontology” as a resource in RDF. The RDF statements are as following: cheap appropriate expensive 100 200 Figure 1. Membership functions in ontology

Basic fuzzy rdf: Seq ontology linguistic values include rdf: Type

rdf_1 rdf_2 rdf_3

Fuzzy concept 1 Fuzzy concept 2 Fuzzy concept 3

membership function membership function membership function

Membership Membership Membership function 1 function 2 function 3

Figure 2. Linguistic variable ontology representation in RDF

(2) Trapezoidal. A trapezoidal shaped curve can be VI. INFORMATION RETRIEVAL FOR E-COMMERCE described by four points, namely: (x1, 0), (x2, 1), (x3,1), In the open and distributed environments of WWW, and (x4, 0). The RDF statements are as following: in order to integrate and reuse information and knowledge in E-commerce, ontology becomes the means to model knowledge for customer [17]-[18] and product [19]-[21]. But the standard ontology is not able to handle fuzzy phenomenon and uncertainty of information and knowledge. In fact, it is sufficient for managers and customers to obtain some message in linguistic values rather than in accurate numeric values, such as customer information, product information and supplier information etc. For instance, the linguistic values for customer income include “low”, “middle”, “high” etc, and linguistic values for product price include “cheap”, 0 “appropriate”, “expensive” etc. These linguistic values have uncertainty and are fuzzy concepts. Using the three-layered fuzzy ontology framework, we construct the ontology structure for customer, product 100 and supplier knowledge shown in Fig. 4, in which the linguistic values are represented formally through fuzzy linguistic variable ontologies. The main fuzzy linguistic variable ontologies are as following: 200 O1= (age, {old, middle-aged, midlife, youth, youngster, adult , …}); O2= (income, {little, low, middle, high, …}); V. FUZZY ONTOLOGY FRAMEWORK O3= (customer type, {new customer, loyalty Combining fuzzy domain ontology with fuzzy customer, gold customer, big customer, lost customer, linguistic variable ontology, we obtain the three-layered switched customer …}); fuzzyontologyframeworkshowninFig.3. O4= (price,{very cheap, cheap, appropriate, The framework comprises the set of concepts and expensive, very expensive}); relations, set of properties and set of fuzzy linguistic O5= (zone of influence, {regional, national, variable ontologies. The relation between concept and international, …}); property is “property of”, and the relation between O6= (quality, {poor, middle, good, very good}); property and fuzzy linguistic variable ontology is “value O7= (delivery time, {very deferred, deferred, on of”, in which property’s value can be either standard data time}); type or linguistic values of fuzzy concepts. The O8= (evaluate grade, { very weak, weak, neutral, framework is the extension of RDF data model “resource- strong, excellent}). property-value”. Since considering the essential semantic There is a lot of semantic relation between fuzzy relationships between fuzzy concepts, the framework concepts. For instance: facilitates the information retrieval at semantic level.

c1 c2 c3 r1 r2 r3

p1 p2 p3 p4 p5 p6

standard data types c: concept O F O F r: relation {boolean,integer, p: property CF {c1,c2,,cn} CF {c1,c2,,cn} float,string, OF:fuzzyontology R={inclusion, intersection, R={inclusion, intersection, symbol, property of disjointness, =,} disjointness, =,} enumeration, instance,class, …} value of F {f1, f2,, fn} F {f1, f2,, fn}

Figure 3. Three-layered fuzzy ontology framework

x “middle-aged” = “midlife”, “old” “adult”, their semantic relation. The part of RDF statements to “middle-aged” “adult”, “youth” “adult”; represent these ontologies is as following: x “gold customer”= “big customer”, “switched customer” “lost customer”; x “very cheap” “cheap” “appropriate” “expensive” “very expensive”; x “very weak” “weak” “neutral” “strong” “excellent”; x ŉ“on time”= {“very deferred” , “deferred”}; reversion (“poor”)= “good”, reversion x (“cheap”)= “expensive”, etc. Fig. 5 shows the RDF graph for linguistic variable ontology O1 which includes a set of fuzzy concepts and

Age (age, {old, middle-aged, youth,…})

Type (customer type, {new customer, loyalty customer, gold Customer customer, …}) Income (income, {little, low, middle, high, …}) purchase property of value of

Price (price,{very cheap, cheap, appropriate, expensive, very expensive}) Product Zone of influence (zone of influence, {regional, national, international, …})

supply Quality (quality, {poor, middle, good, very good})

Supplier Delivery time (delivery time, {very deferred, deferred, on time}) property of value of

Evaluate (evaluate grade, { very weak, weak, neutral, strong, grade excellent})

Figure 4. Ontology structure for customer, product and supplier knowledge (portion)

ex: age rdf: Seq ex: linguistic values include rdf: Type

rdf_1 rdf_2 rdf_3 rdf_4

ex: ex: youngster ex: youth ex: midlife ex: old ex: ex:

ex: = ex: ex: ex:

ex: middle-aged ex: adult ex:

Figure 5. An example of RDF graph for linguistic variable ontology

Query Retrieval engine request Inference machine Domain ontology Furthermore, we build the information retrieval system shown in Fig.6. Since the process for information Query retrieval is based on the knowledge ontology, the result Metadata semantic and concept research can be achieved. Data source Especially, using linguistic value of fuzzy concept, we can construct the research pattern such as: SELECT Figure 6. Information retrieval system instance of concept FROM Data source WHERE (property of concept) Using the “complement relation” defined in fuzzy “Linguistic value of fuzzy concept”, in which the linguistic variable ontology: ŉ“on time”= {“very comparison operators includes: equal comparison (=), deferred”, “deferred”}, when retrieving the information unequal (), less than or equal () and greater than or about supplier by the statement: SELECT Supplier (name, equal () etc. address, phone,…) FROM Data source WHERE For instance, we can retrieve “product” information Supplier.delivery timeĮ“on time”, we can transform the through “price” of property, using the search statement search statement to: SELECT Supplier ( name, address, such as: SELECT Product ( name, brand, price, …) phone, …) FROM Data source WHERE Supplier.delivery FROM Data source WHERE Product.price time = “very deferred” or Supplier.delivery = “ expensive”. The standard ontology and other fuzzy “deferred”. ontology are not able to handle the search condition at Using the “reversion relation” defined in fuzzy semantic level, which includes fuzzy concept and linguistic variable ontology: reversion (“cheap”) = semantic relation between them. “expensive”, when retrieving the information about Using the “order relation” defined in fuzzy linguistic product by the statement: SELECT Product ( name, brand, variable ontology : “very cheap” “cheap” price, …) FROM Data source WHERE Product.price = “appropriate” “expensive” “very expensive” , we can REVERSION (“cheap”), we can transform the search transform the search statement to: SELECT Product statement to: SELECT Product ( name, brand, price, …) ( name, brand, price, …) FROM Data source WHERE FROM Data source WHERE Product.price = Product.price = “very cheap” or Product.price = “expensive”. “cheap” or Product.price = “appropriate” or Product.price = “expensive”, in which every sub- VII. CONCLUSION condition is ordinary and can be completed easily in SQL People can obtain information from data resources by engine of DBMS. semantic querying based on ontology. To achieve fuzzy When retrieving the information about gold customer semantic retrieval in E-Commerce, this paper has by the statement: SELECT Customer (name, age, presented information retrieval system based on fuzzy income,…) FROM Data source WHERE ontology framework. The framework includes three parts: Customer.type=“gold customer”, we can obtain the concepts, properties of concepts and values of properties, information about big customer using equivalence in which property’s value can be either standard data type relation: “gold customer”= “big customer”. or linguistic values of fuzzy concepts. The framework is When retrieving the information about lost customer the extension of RDF data model “resource-property- by the statement: SELECT Customer (name, age, value”, which is the current standard for establishing income,…) FROM Data source WHERE semantic interoperability on the Semantic Web. The Customer.type=“lost customer”, we can obtain the semantic query expansions have been constructed by information about switched customer using inclusion order relation, equivalence relation, inclusion relation, relation: “switched customer” “lost customer”. In the complement relation between fuzzy concepts defined in same way, using inclusion relation: “old” “adult”, fuzzy linguistic variable ontologies, which facilitates the “middle-aged” “adult” and “youth” “adult” and information retrieval at semantic level. equivalence relation: “middle-aged” = “midlife”, we can Our further researches lay on the semantic query transform the search statement: SELECT Customer (name, expansion using complex fuzzy semantic relations in age, income,…) FROM Data source WHERE RDF query language such as RDQL, RQL, SeRQL. Customer.age=“adult” to: SELECT Customer (name, age, income,…) FROM Data source WHERE ACKNOWLEDGMENT Customer.age=“adult” or Customer.age = “old” or Customer.age = “middle-aged” or Customer.age = This work was supported in part by the Research “midlife” or Customer.age = “youth”. Project of the Educational Department of Liaoning

Province (Leading Laboratory Project) under Grant ontology framework,” in: Proceedings of the 2006 20060083. IEEE/WIC/ACM International Conference on Web Intelligence, Hong Kong, 2006, pp.397-403. REFERENCES [13] Calegari Silvia, Ciucci Davide, “Fuzzy ontology and fuzzy-OWL in the KAON project,” in: Proceedings of [1] D. Fensel, F. van Harmelen, I. Horrocks, D.L.McGuinness, 2007 IEEE International Conference on Fuzzy Systems and P.F. Patel-Schneider, “OIL: an ontology infrastructure Conference, London, UK, 2007, pp.1-6. for the semantic web,” IEEE Intelligent Systems, vol.16, [14] K. Atanassov, “Operators Over Interval-valued no.2, 2001, pp. 38-45. Intuitionistic Fuzzy Sets,” Fuzzy Sets and Systems, vol.64, [2] T. Berners-Lee, “Semantic web road map, W3C design no.2, 1994, pp.159-174. issues”, http://www.w3.org/DesignIssues/Semantic.html, [15] T. R. Gruber, “A Translation Approach to Portable 1998. Ontology Specification,” Knowledge Acquisition,vol.5, [3] P. Castells, M. Fernandez, D. Vallet, “An adaptation of the no.2, pp. 199-220. vector-space model for ontology-based information [16] John W.T. Lee, Alex K.S. Wong, “Information retrieval retrieval,” IEEE Transactions on Knowledge and Data based on semantic query on RDF annotated resources,” in: Engineering, vol.19, no. 2, 2007, pp. 261-272. Proceedings of the 2004 IEEE International Conference on [4] J. Li, J. Y. Song, H. Zhong, “Ontology-based query Systems, Man and Cybernetics, 2004, pp. 3220-3225. division and reformulation for heterogeneous information [17]X.W.Lu,F.Jiang,L.W.Hou,“Customerfeatures integration,” Journal of Software, vol.18, no.10, 2007, pp. extraction based on customer ontology,” Computer 2495í2506. (in Chinese) Engineering, vol.31, no.5, 2005, pp. 31-33. (in Chinese) [5] L. A. Zadeh, J. Kacprzyk, Fuzzy logic for the management [18] Ma Huinan, Wu Jiangning, Dang Yanzhong, “Ontology- of uncertainty, John Wiley & Sons, Inc, 1992. driven information retrieval system for customers' [6] D. H. Widyantoro, J. Yen, “A fuzzy ontology-based complaints of mobile communication services,” in: abstract search engine and its user studies,” in: Proceedings of 2006 International Conference on Proceedings of the 10th IEEE International Conference on Management Science and Engineering, Lille, 2006, pp. 30- Fuzzy Systems, Melbourne, Australia, 2001, pp. 1291- 35. 1294. [19] Hu Yujie, Li Shanping, Guo Ming, “Ontology Based [7] D. Parry, “A fuzzy ontology for medical document Product Knowledge Representation,” Journal of Computer retrieval,” in: Proceedings of the Second Workshop on Aided Design & Computer Graphics, vol. 15, no.12, 2003, Australasian Information Security, Data Mining and Web pp. 1531-1537. (in Chinese) Intelligence, and Software Internationalization,Dunedin, [20] Fu Xiangjun, Li Shanpin, Guo Ming, “Ontology New Zealand, 2004, pp. 121- 126. Knowledge Representation for Product Data Model,” [8] Raymond Y.K. Lau, “Fuzzy domain ontology discovery Journal of Computer Aided Design & Computer Graphics, for business knowledge management,” IEEE Intelligent vol.17, no.3, 2005, pp. 570-577. (in Chinese) Informatics Bulletin, vol.8, no.1, 2007, pp. 29-41. [21] Fu Xiangjun, Li Shanping, Guo Ming, “Rules for Ontology [9] C.S.Lee,Z.W.Jian,andL.K.Huang,“Afuzzyontology Level Expression in Product Data,” Journal of Computer and its application to news summarization,” IEEE Aided Design & Computer Graphics, vol.17, no.6, 2005, Transactions on Systems, Man and Cybernetics (Part B), pp. 1159-1166. (in Chinese) vol.35, no.5, 2005, pp. 859- 880. [10] Q. T. Tho, S. C. Hui, A. C. M. Fong, and T. H. Cao, “Automatic fuzzy ontology generation for semantic web,” IEEE Transactions on Knowledge and Data Engineering, vol.18, no.6, 2006, pp. 842- 856. [11] Muhammad Abulaish, Lipika Dey, “A fuzzy ontology generation framework for handling uncertainties and Jun Zhai was born in Changchun, P.R. China, on Sep. 19, nonuniformity in domain knowledge description,” in: 1969. He graduated from the Harbin Institute of Technology, Proceedings of 2007 International Conference on Ph.D, and studied at the Dalian Maritime University, associate Computing: Theory and Applications, Kolkata, 2007, pp. professor. He is now a post-doctoral scholar and his research 287-293. interests include knowledge management and its application in [12] Muhammad Abulaish, Lipika Dey, “Interoperability Electronic Commerce and transport system. among distributed overlapping ontologies – a fuzzy