J.S. Ribeiro, K.A. Kaufman, and L. Kerschberg. Knowledge discovery from multiple databases. In Knowlege Discovery and Datamining, Menlo Park, CA (1995).  Home/Search   Document Not in Database   Summary   Related Articles   Check

....in different databases, and transferring them (or carefully selected subsets) to the data mining platform. If network access were provided to these data, through SQL servers or tailored data mining servers, an inductive algorithm could query the remote databases as necessary during data mining. Ribeiro, Kaufman and Kersberg (1995) describe a method for performing knowledge discovery across multiple databases by using foreign key values to augment tables. Specifically, they propose tracing through multiple databases following the foreign keys, and learning individual knowledge segments for each database. The WoRLD system ....

Ribeiro, J., K. Kaufmann, and L. Kerschberg (1995). Knowledge discovery from multiple databases. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Menlo Park, CA, pp. 240--245. AAAI Press.

....to express larger queries. After the analysis of the request has been performed, a thesaurus [Jing94] defined by the experts will be used in order to retrieve the the relevant attributes unsuspected by the user. Further work is to integrate paradigms from knowledge representation and discovery [Ribeiro95] . 4 Conclusion This article presented the main problems related to the design of a cartography tool integrating multiple and various information sources and we propose a cross fertilization method based on the usage of an hyper document. We mainly deal with the selection of valuable database ....

Ribeiro, J. and Kaufman, K. and Kerschberg, L. Knowledge discovery from multiple databases, Proc. IASTED/ISMM Int. Conf. on Intelligent Information Management Systems, June 1995.

....why this is an important question. First, similar to the data partitioning described above, concurrent analysis of different relations may give additional speedups. Second, a database of interest may be accessible over the network, but not practically transferrable. Ribeiro, Kaufman and Kersberg [65] describe a method for performing knowledge discovery across multiple databases by using foreign key values to augment tables. Specifically, they propose to trace through multiple databases following the foreign keys, and learn individual knowledge segments for each database. The WoRLD system [6] ....

Ribeiro, J.S., Kaufmann, K.A., and Kerschberg, L. (1995). Knowledge Discovery from Multiple Databases. In The Proc. of The First Intl. Conf. on Knowledge Discovery and Data Mining (KDD96) , Menlo Park, CA: AAAI Press, pp: 240-245.

....relevant tables and transferring them (or carefully selected subsets) to the data mining platform. If network access were provided to these data, either as an SQL server, or a tailored data mining server, an inductive algorithm could query the remote database as necessary during data mining. Ribeiro, Kaufman and Kersberg (1995) describe a method for performing knowledge discovery across multiple databases by using foreign key values to augment tables. Specifically, they propose tracing through multiple databases following the foreign keys, and learning individual knowledge segments for each database. The WoRLD system ....

Ribeiro, J., K. Kaufmann, and L. Kerschberg (1995). Knowledge discovery from multiple databases. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Menlo Park, CA, pp. 240--245. AAAI Press.

....on synthetic data sets, and then replicate two real world successes of automated discovery. 1 INTRODUCTION Inductive machine learning offers methods for discovering new knowledge from business, medical, and scientific databases. Although the need to learn across multiple tables has been realized [17], most inductive learning and data mining techniques assume that all the relevant information for discovery has been gathered and assembled into a single table or database. With multiple tables and multiple databases it is possible to combine features from several perspectives and thus move beyond ....

....and databases. This requires a domain expert to determine which databases and which fields within them are relevant. Combining multiple databases also creates scaling problems for discovery programs. Even recent work that has begun to address the problem of learning across multiple databases [17] requires that the databases reside on the same machine. In practice, useful databases may exist in remote locations in an organization, or across the Internet, and in either case, their existence may not be known to the persons initiating an inquiry. The WoRLD system described in this paper is an ....

Ribeiro, J.; Kaufman, K.; Kerschberg, L. 1995. Knowledge discovery from multiple databases. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining. AAAI Press.

....the variable valued logic language VL1 (Michalski, 1973) and Annotated Predicate Calculus (Michalski, 1983) has a very high descriptive power as well as cognitive simplicity. This schema leads to a natural division of a rule base into subunits, each describable by a set of pertinent statistics (Ribeiro, Kaufman and Kerschberg, 1995). An example of this structure is shown in Figure 3. The application of an operator (an individual learning Session) is associated with an input data set, the time at which it was run, and a set of output Class Rulesets. Each of these Class Rulesets can be associated with the class of objects it ....

Ribeiro, J., Kaufman, K. and Kerschberg, L., "Knowledge Discovery from Multiple Databases," Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Montreal PQ, pp. 240-245, 1995.

.... i.e. learning from examples in which the values of some attributes are unknown (e.g. Dontas, 1988; Lakshminarayan et al., 1996) Learning from distributed data, i.e. learning from separate collections of data that must be brought together if the patterns within them are to be exposed (e.g. Ribeiro, Kaufman and Kerschberg, 1995). Learning drifting or evolving concepts, i.e. learning concepts that are not stable but changing over time, randomly or in a certain general direction. For example, the area of interest of a user is often an evolving concept (e.g. Widmer and Kubat, 1996) Learning concepts from data ....

Ribeiro, J., Kaufman, K. and Kerschberg, L., "Knowledge Discovery from Multiple Databases," Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Montreal PQ, pp. 240-245, 1995.

.... from incomplete data, i.e. learning from examples in which the values of some attributes are unknown (e.g. Don88] LHGS96] Learning from distributed data, i.e. learning from separate collections of data that must be brought together if the patterns within them are to be exposed (e.g. [RKK95]) Learning drifting or evolving concepts, i.e. learning concepts that are not stable but changing over time, randomly or in a certain general direction. For example, the area of interest of a user is often an evolving concept (e.g. WK96] Learning concepts from data arriving over time, ....

Ribeiro, J.S., Kaufman, K.A. and Kerschberg, L. Knowledge Discovery From Multiple Databases. Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Montreal, PQ, pp. 240-245, 1995.

....of the problem and the nature of a likely solution and report it immediately. INLEN currently examines only one relational table at a time. An ongoing research effort is to develop methods for discovering knowledge that is not stored in one location, but rather distributed among several databases (Ribeiro et al., 1995). One approach avoids combining the data from the different sources, instead developing separate knowledge bases and deriving rule based attributes in order to search for commonalities between the discovered knowledge and the other databases. A fourth research effort applicable to INLEN is the ....

Ribeiro, J.S., Kaufman, K.A. and Kerschberg, L. (1995). Knowledge Discovery From Multiple Databases.

No context found.

J.S. Ribeiro, K.A. Kaufman, and L. Kerschberg. Knowledge discovery from multiple databases. In Knowlege Discovery and Datamining, Menlo Park, CA (1995).

No context found.

J. Ribeiro, K. Kaufman, and L. Kerschberg, Knowledge discovery from multiple databases. In: Proceedings of KDD95. 1995: 240-245.

No context found.

J. Ribeiro, K. Kaufman, and L. Kerschberg, Knowledge discovery from multiple databases. In: Proceedings of the First International Conference on Knowledge Discovery and Data Mining (KDD-95), Montreal, Canada, AAAI Press, August 20-21, 1995: 240-245.

No context found.

J. Ribeiro, K. Kaufman, and L. Kerschberg, Knowledge discovery from multiple databases. In: Proceedings of the First International Conference on Knowledge Discovery and Data Mining (KDD-95), Montreal, Canada, AAAI Press, August 20-21, 1995: 240-245.

No context found.

J. Ribeiro, K. Kaufman, and L. Kerschberg, Knowledge discovery from multiple databases. In: Proceedings of KDD95. 1995: 240-245.

No context found.

J.S. Ribeiro, K.A. Kaufman, and L. Kerschberg. Knowledge discovery from multiple databases. In Knowlege Discovery and Datamining, Menlo Park, CA (1995).

No context found.

Ribeiro, J.S., Kaufman, K.A., and Kerschberg, L. Knowledge discovery from multiple databases, Proceedings of KDD'95, 1995

No context found.

Ribeiro, J.S., Kaufman, K.A., and Kerschberg, L. Knowledge discovery from multiple databases, In Proceedings of the First International Conference on Knowledge Discovery & Data Mining (KDD'95), 240-245, 1995

No context found.

Ribeiro, J.S., Kaufmann, K., and Kerschberg, L. (1995). Knowledge Discovery from Multiple Databases, In Proc. of the 1 st Int'l Conf. On Knowledge Discovery and Data Mining, Quebec, Montreal.

No context found.

Ribeiro, J.S., Kaufmann, K., and Kerschberg, L. Knowledge Discovery from Multiple Databases. In Proc. of the 1 st Int'l Conf. On Knowledge Discovery and Data Mining, Quebec, Montreal, 1995.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC