Chun-Nan Hsu and Craig A. Knoblock. Estimating the robustness of discovered knowledge. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Menlo Park, CA, 1995. AAAI Press.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....step studies how to analyze the discovered patterns and use them, together with the domain knowledge to create new KDD strategies. In this step human knowledge and common sense reasoning are still irreplaceable. Some of the tasks that should be attacked at this step are: ffl Pattern evaluation [53, 103, 19]. The purpose of this task is to analyze how interesting a discovered pattern can be in order to avoid showing the user a large set of uninteresting patterns. ffl Conflict resolution. When there is a conflict between the domain knowledge and the discovered knowledge, we are facing an interesting ....

Chun-Nan Hsu and Craig A. Knoblock. Estimating the Robustness of Discovered Knowledge. In U.M. Fayyad and R. Uthurusamy, editors, Proceedings of the First International Conference on Knowledge Discovery and Data Mining, 1995.

....component exploits learned knowledge about the contents of databases to perform semantic query optimization. We have extended semantic query optimization techniques to support multidatabase queries and have developed an approach to learn the rules for the optimization process. See [10] 11] [12], 13] for details. ffl Execution Finally, the fourth component executes the optimized query plan. SIMS executes queries against the appropriate information sources (doing so in parallel when possible) transfers data, constructs a response to the user, and returns it. An execution failure will ....

Chun-Nan Hsu and Craig A. Knoblock. Estimating the robustness of discovered knowledge. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Menlo Park, CA, 1995. AAAI Press.

....third component exploits learned knowledge about the contents of databases to perform semantic query optimization. We have extended semantic query optimization techniques to support multidatabase queries and have developed an approach to learn the rules for the optimization process. See [9] 10] [11], 12] for details. ffl Execution Finally, the fourth component executes the optimized query plan. SIMS executes queries against the appropriate information sources (doing so in parallel when possible) transfers data, constructs a response to the user, and returns it. An execution failure will ....

Chun-Nan Hsu and Craig A. Knoblock. Estimating the robustness of discovered knowledge. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Montreal, Canada, 1995.

....queries and transactions. The goal of this research is to develop a general learning approach that maximizes the net utility of semantic rules throughout the life span of a information system. To achieve this goal, I have developed an approach to the estimation of the robustness of semantic rules [Hsu and Knoblock, 1995]. This estimation approach allows the learner to determine the degree of the robustness of a semantic rule and use the results to guide the search for robust rules. This estimation approach will be described in detail in Chapter 2. Robustness is different from predictive accuracy in rule ....

Chun-Nan Hsu and Craig A. Knoblock. Estimating the robustness of discovered knowledge. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining(KDD-95), Menlo Park, CA, 1995. AAAI Press.

....semantic query optimization is to search for the least expensive query in the space of semantically equivalent queries. The transformation from one query to another is done through logical inference using database abstractions, the abstracted knowledge of the contents of relevant databases. See [ Hsu and Knoblock, 1994; 1995; 1996 ] for an explanation of how rules like these are automatically learned. The database abstractions describe the databases in terms of the set of closed formulas of first order logic. These formulas describe the database in the sense that they are true with regard to all instances in the ....

Chun-Nan Hsu and Craig A. Knoblock. Estimating the robustness of discovered knowledge. In Proceedings of the First International Conference on Knowledge Discovery and Data Mining, Montreal, Canada, 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