[Home] [Current Edition] [Compendium] [Forum] [Web Archive]
[Email Archive] [Guestbook] [Subscribe] [Advertising Rates]

Knowledge Grid: Enabling Intelligent Internet Marketing

By Dr Hai Zhuge, Chinese Academy of Sciences

Email: zhuge@ict.ac.cn

Dr Hai ZHUGE is a full professor at the Institute of Computing Technology, Chinese Academy of Sciences. His research interests include: e-commerce model and system, knowledge management, next-generation web, and co-operative process model. He is now leading twenty team members working on the China Knowledge Grid project VEGA-KG. Recently, he has proposed the Resource Space Model RSM and the supported methods and tools for uniformly sharing and managing e-resources across the Internet for the first time. He is the first author of one book and over 50 papers appeared mainly in leading international conferences and the following international journals: IEEE Transactions on Systems, Man, and Cybernetics; Information and Management; Decision Support Systems; Journal of Systems and Software; Expert Systems with Applications, Knowledge-based Systems; Information and Software Technology; and, Journal of Software. He is playing the editorial roles in several international journals.


Efficient Internet market should support intelligent business behavior, but current web technology is unable to support knowledge sharing and management. Enabling intelligence is one of the targets of the next-generation web. This paper introduces a Knowledge Grid platform VEGA-KG, where VEGA stands for versatile resources, Enabling intelligence, Global uniformity and Autonomous control. It enables the uniform sharing and management of knowledge resources across the Internet. The VEGA-KG includes two major components: a resource space model that uniformly organizes information, knowledge, and service resources in normal forms; and an operable knowledge browser that enables users to conveniently locate and manage resources by using and easy-to-use interface. This platform enables geographically distributed participants with Internet access to share business knowledge.


    In contrast to the rapid expansion of business information, Internet users still lack effective means to publish, organize, share, and manage knowledge resources across the Internet. Thus, agents or people in virtual markets are isolated from each other at the knowledge level, and must do business from scratch relying only on their own knowledge. Knowledge management plays an important role in promoting innovation and productivity of organizations [1, 3, 4, 6, 10]. Knowledge management in Internet markets should enable market participants to store their knowledge at any time when they have generated some useful knowledge, and to easily retrieve desired knowledge from the repositories distributed on the Internet. In this way, knowledge resources in Internet markets can rapidly accumulate and evolve as the common knowledge assets of the whole Internet community grow with the expansion of the Internet market participants. The Semantic Web (http://www.semanticweb.org) and the Grid (http://www.gridforum.org) are two approaches towards the next-generation web. The current research on the Semantic Web focuses on ontology and markup languages like XML, RDF and DAML [2]. A Grid can be regarded as an integrated platform that enables the controlled sharing of versatile resources. A generic Grid should have four characteristics: network ability, interoperability, composition ability, and semantic completeness [5]. VEGA is a Grid project launched by the Chinese Academy of Sciences with four goals: provide versatile service, enable intelligence, create a global standard, and provide users with autonomous control.


    2.1 Resource Space

    The Resource Space Model is for uniformly specifying information, knowledge, and service resources in an n-dimensional space. It has the following three distinguished characteristics:

    1) Normalized coordinate system. The resource space model provides three normal forms to normalize a resource space so as to guarantee the correctness of resource operations just as the relational data model.

    2) Provide both the local view and the universal resource view. The resource space model enables users to choose to operate resources in either the local resource view or the universal resource view by joining different local resource spaces into one resource space.

    3) Support the management of the structured or semi-structured resources. Templates are used to uniformly represent versatile resources. In the resource space model, a resource operation language is provided for uniformly operating resources, a set of criteria is set to help designers create a good resource space, and a development method is provided to guide designers who are developing a resource space. The main difference between the RSM and the relational data model includes six aspects: the foundation, the managed objects, the data model, the normalization basis, the operation feature, and the interchange basis. Further investigation shows that the RSM is also suitable for managing relational tables. The RSM can be used in many other application fields like digital library and component repositories [7, 8].

    2.2 Knowledge Space

    A knowledge space is a special case of the resource space, it has a three-dimensions: knowledge-level, knowledge-category, and location.

    1) Knowledge Level. It includes four co-ordinates: concept, axiom, rule, and method.

    2) Knowledge Category. Knowledge category reflects the disciplines of the knowledge.

    3) Location. Universal Knowledge Location (UKL). The format of the UKL is: URL/[GroupName/]UserName/[attribute]/[x,y/].

    2.3 Knowledge Browser

    Knowledge browser is an easy-to-use and operable browser, which is responsible for:

    1) Locating knowledge resources in knowledge space by determining its coordinates;

    2) Selecting suitable operations and setting the parameters;

    3) Delivering the operation to the execution engine; and,

    4) Receiving and showing the operation results.

    2.4 Operation Function Interfaces

    The VEGA-KG also provides operation function interface for automatic agents to use knowledge operations without human operations. All the functions are specified in XML format. Agents can subscribe functions from VEGA-KG, which carries out search and then provide the agents with the suitable functions.


    The VEGA-KG provides the Internet marketing participants the platform to effectively share and manage knowledge resources across the Internet. Different from the other Grid model, VEGA-KG is based on the resource space model RSM and the Semantic Web, which enables knowledge resources to be machine-understandable so as to support intelligent business process. The first version of the VEGA-KG has been implemented, and is available for use at http://kg.ict.ac.cn. Ongoing work is to develop a Process Grid that is above the Knowledge Grid and can manage versatile business processes on the web based on the proposed resource space model and the workflow technology [9], and to carry out e-business experiments on the VEGA-KG platform.


    • R.Dieng, Knowledge Management and the Internet, IEEE Intelligent Systems, May/June, 2000, pp.14-17.

    • M.Klein, XML, RDF, and Relatives, IEEE Internet Computing, March/April, 2001, pp.26-28.

    • P.Martin and P.W.Eklund, Knowledge Retrieval and the World Wide Web, IEEE Intelligent Systems, May/June, pp. 18-25, 2000.

    • H.Zhuge, J.Ma, and X.Shi, Analogy and Abstract in Cognitive Space: A Software Process Model, Information and Software Technology, vol.39, pp.463-468, 1997.

    • H.Zhuge, A Knowledge Grid Model and Platform for Global Knowledge Sharing, Expert Systems with Applications, vol.22, no.4, 2002.

    • H.Zhuge, A Knowledge Flow Model for Peer-to-Peer Team Knowledge Sharing and Management, Expert Systems with Applications, vol.23, no.1, 2002.

    • H.Zhuge, A Problem-oriented and Rule-based Component Repository, Journal of Systems and Software, vol.50, pp. 201-208, 2000.

    • H.Zhuge, Inheritance Rules for Flexible Model Retrieval, Decision Support Systems, vol.22, No.4, pp.379-390, 1998.

    • H.Zhuge, T.YCheung, and H.K.Pung, A Timed Workflow Process Model, Journal of Systems and Software, vol.55, pp.231-243, 2001.

    • H.Zhuge and X. Shi, Communication Cost of Cognitive Co-operation for Team Development, Journal of Systems and Software, vol.57, issue 3, pp.227-233, 2001.