RETRIEVAL SYSTEMS
Version 2 (Separate print control knowledge from domain knowledge): %Print control knowledge
5.9 KNOWLEDGE MANAGEMENT AND ONTOLOGIES In this chapter, started from data retrieval systems, we extended our
5.9.3 DATA AND KNOWLEDGE MANAGEMENT ONTOLOGIES
Since the knowledge available in a knowledge management system comes from various sources and takes various formats, it is a big challenge to use and reuse such acquired knowledge in an integrated manner. Similarly, in distributed database management systems, multi-database systems or data warehouses, ontologies also play an important role. To deal with this challenge, we have to consider issues related to ontology, which is explicit, knowledge-based specifications of conceptualizations. These specifications typically describe a taxonomy of the tasks that define the knowledge. Within the context of knowledge management systems, ontology is the specifications of discourse in the form a shared vocabulary. Ontology thus plays an important role of integrated use of knowledge in an organization.
There are significant advantages of using ontology in knowledge management. First of all, ontology defines the scope of group discussions needed by knowledge management systems and serves as the common language for collaboration. As a consequence, ontology also facilitates reusability of artifacts achieved in knowledge management systems. In addition, ontology provides more focused search capabilities needed in organizations, filters substantial amounts of information, and directs the information of interest to the appropriate source. In order to select an appropriate ontology, a number of factors should be considered. More discussion on knowledge management can be found in [O'Leary 1998a, 1998b, Borghoff and Pareschi 1997].
A collection of papers on ontologies can be found in [Swartout and Tate 1999] and a collection of papers on ontologies in distributed databases can be found in [Bougurettaya 1999]. The American Heritage Dictionary defines "ontology" as "the branch of metaphysics that deals with the nature of being. " (Metaphysics refers to the branch of philosophy that systematically investigages first causes and the nature of ultimate reality.) The term has recently been adopted by the computational intelligence community to refer to a set of concepts or terms that can be used to describe some area of knowledge or build a representation of it. An ontology can be either very high level (consisting of concepts that organize the upper parts of a knowledge base) or domain specific. An ontology provides the basic structure or armature around which a knowledge base can be built. The distinction between an ontology and a knowledge base lies in that an ontology provides a set of concepts and terms for describing some domain, while a knowledge base uses those terms to represent what is true in that domain. Interest in ontologies is largely due to
reusing or sharing knowledge across systems [Swartout and Tate 1999]. One key impediment to sharing knowledge is that diferent systems use different concepts that terms for describing domains. Ontologies will fundamenally change the way in which systems are constructed. Of particular interest is the issue of the use of databases over the Web. Because of the sheer size of the Web, the data volume is steadily becoming larger, and the information space is increasingly dynamic. In light of these developments, one emerging area that holds promise to define a common representation and understanding is the use of ontologies in databases, which have drawn from computational intelligence, linguistics and philosophy [Bougettaya 1999].
SUMMARY
In this rather long chapter we discussed various kinds of retrieval systems: database retrieval, information retrieval and knowledge retrieval. A good understanding on the similarities and differences of these systems is crucial for the integrated use of these systems for decision support. Materials presented in this chapter, along with those presented in the next chapter, will form the core of database and knowledge-based systems. A collection of recent research papers [Yang 1999] examine recent development in intelligent information retrieval, including searching, filtering and navigating on the Web; multimedia information retrieval; and the incorporation of machine learning techniques into intelligent retrieval (see Chapter 10 for a discussion on machine learning). Some advanced issues related to intelligent retrieval, including reasoning through extended retrieval, as well as integrated retrieval involving creativity, will be further discussed in Chapters 7 to 9.
SELF-EXAMINATION QUESTIONS
1. Make your examples to illustrate how to integrate information retrieval and database retrieval, and discuss some advantages as well as some issues must be considered.
2. Suppose you heard from the news report that Miami is declared as the capital of the United States.
(a) Indicate all the possible implications you can make from this news.
(b) Suppose you want to write a Prolog program to produce these implications. Discuss what kind of facts and assumptions should be stored in the knowledge base.
(c) Instead of writing a Prolog program, suppose you are asked to accomplish the same task indicated in (b) by developing a knowledge-based system. The knowledge-based system will not store any facts; rather, it is to be integrated with a database
management system to retrieve all the data needed (such as geographical information) for reasoning. Discuss some important issues must be considered in developing such a system.
3. Consider the issue of handling duplicates in RA and SQL. (A tuple is a duplicate of another one if they are identical.) Answer the following questions in regard to operators used in RA and SQL:
(a) Which operators retain duplicates?
(b) Which operators automatically eliminate duplicates? (c) When and why should duplicates be retained? (d) Are duplicate explicitly removed or retained?
4. Consider the simple expert system example discussed in Section 5.8.7.3. Design a system-user conversation under which the right subtree at the node diagnose(X) will not be searched at all. Is it possible to prune the left subtree at the same node?
REFERENCES
Borghoff, U. M. and Pareschi, R., Information technology for knowledge management. Journal of Universal Computer Science, 3(8), 835-842, 1997. Bouguettaya, A. (guest ed.), Ontologies and databases (special issue), Distributed and Parallel Databases, 7(1), 5-98, 1999.
Brachman, R. J. and Levesque, H. J., What makes a knowledge base knowledgeable? A view of databases from the knowledge level, in Kerschberg, L. (ed.), Expert Database Systems, 69-78, 1986.
Chandrasekaran, B., Generic tasks in knowledge-based reasoning: High- level building blocks for expert system design, IEEE Expert, 1(3), 23-29, June 1986.
Chaudhuri, S. (ed.), Special issue on databases and the World Wide Web, Data Engineering Bulletin, pp. 3-52, 21(2), 1998.
Chen, Z., Enhancing database management to knowledge base management: the role of information retrieval technology, Information Processing and Management, 30(3) 419-435, 1994.
Dean, T., Allen, J. and Aloimonos, Y., Artificial Intelligence: Theory and Practice, Benjamin/Cummings, Redwood City, CA, 1995.
Eich, M. (ed.), Special section on mian memory databases, IEEE Transactions on Knowledge and Data Engineering, 4(6), 507-571, 1992.
Floreskcu, D., Levy, A. and Mendelzon, A., Database techniques for the World Wide Web: A survey, SIGMOD Record, 27(3), 59-74, Sept. 1998. Frakes, W. B. and Baeza-Yates, R. (eds.), Information Retrieval: Data Structures and Algorithms, Prentice-Hall, Englewood Cliffs, NJ, 1992.
Freundlich, Y., Knowledge Bases and Databases: Converging Technologies, Diverging Interests, IEEE Computer, 23(11), 51-58, 1990.
Garcia-Molina, H.L., Labio, W. J., Wiener, J. L. and Zhuge, Y., Distributed and parallel computing issues in data warehousing, Proceedings of ACM Principles of Distributed Computing Conference, 1999.
Giarratano, J. and Riley, G., Expert Systems: Principles and Programming (3rd ed.), PWS Publishing Co., Boston, 1998.
Harinarayan, V., Issues in interactive aggregation, Data Eng. Bulletin, 20(1), 12-18, 1997.
Inmon,W. H. Building the Data Warehouse. John Wiley, New York, 1996. Kimball, R., The Data Warehouse Toolkit, Wiley, New York, 1996.
Korfhage, R., Information Storage and Retrieval, John Wiley, New York, 1997.
McDermott, J., Preliminary steps toward a taxonomy of problem-solving methods, Chap. 8 in Marcus, S. (ed.), Automating Knowledge Acqusition for Knowledge Based Systems, pp. 120-146, Kluwer, Boston, 1988.
O'Leary, D. E., Knowledge-management systems: Converting and connecting, IEEE Intelligent Systems, 30-33, May/June, 1998a.
O'Leary, D. E. Using AI in knowledge management: Knowledge bases and ontologies, IEEE Intelligent Systems, pp. 34-39, May/June, 1998b.
Ramakrishnan, R., Database Management Systems, WCB McGraw-Hill, Boston, 1998.
Schank, R. C., Issues for psychology, AI, and education: a review of Newell's Unified Theories of Cognition, Artificial intelligence, 59(1/2), 375- 388, 1993.
Silberschatz, A., Korth, H. F. and Sudarshan, S., Database System Concepts (3rd ed.), McGraw-Hill, New York, 1996.
Sparck-Jones, K., Willett, P. and Larson, R. ( e d s . ) , Readings in Information Retrieval, Morgan Kaufman, San Mateo, CA, 1997.
Swartout, W. and Tate, R. (guest eds.), Ontologies, IEEE Intelligent Systems & Their Applications, special issue papers appearing in 14(1), 18-54, 14(2), 63-80, 14(3), 73-79, 14(4), 79-85, 1999.
Ullman, D. D. and Widom, J., A First Course in Database Systems, Prentice Hall, Upper Saddle River, NJ, 1997.
Wiederhold, G., Knowledge and database management, IEEE Computer, 17(1), 63-73, 1984.
Yang, Y. (guest ed.), Intelligent Information Retrieval, IEEE Intelligent System & Their Applications, 14(4), 30-69, 1999.