Traditionally, high-resolution geographic data are managed in a base database and coarser resolution databases are produced in an often costly batch generalization process. Kilpeläinen and Sarjakoski (1995) introduced the concept of an incremental generalization environment where previously generalized databases are updated in an incremental way based on the changes to the base database. The pa...

PANIC (Parallelism And Neural networks In Classifier systems) is a parallel system to evolve behavioral strategies codified by sets of rules. It integrates several adaptive techniques and computational paradigms, such as genetic algorithms, neural networks, temporal difference methods and classifier systems, to define a powerful and robust learning system. To allocate credit to rules, we propos...

We describe the constraint model transformation system Cadmium, focussing on its application. We discuss how we use it to reduce solver-independent models in the medium-level modelling language MiniZinc into equivalent models for finite domain constraint solvers and integer linear programming solvers.

ARKref is a tool for noun phrase coreference that is based on the systemdescribed byHaghighi and Klein (2009) (which was never publicly released). It was originally written in 2009. At the time of writing, the last released version was in March 2011. This document describes that version, which is open-source and publicly available at http://www.ark.cs.cmu.edu/ARKref.1 ARKref is a deterministic,...

In this paper, a ‘market trading’ technique is integrated with the techniques of rule discovery and refinement for data mining. A classifier system-inspired model, the market-based rule learning (MBRL) system is proposed and its capability of evolving and refining rules is investigated. Experimental results indicate that the MBRL system is a potentially useful additional tool that can be used t...

This paper reports a research work to address the problem of the large number of alerts generated by the detectors in an intrusion detection system. Some of these alerts are redundant and have to be aggregated; others may follow a certain attack pattern that should be correlated. Generally, this operation is referred to as alert correlation. A more detailed explanation of the alert correlation ...

.......................................................................................................................... i

.A rule-based system to make illfelences about ~CTU~VPUI images is int reduced. Given a digitized document ‘mage, the JVS tern controls the analysis of the document, and identifLes al! the diJerent printed regions in the document image. I,ogical “blocks” of info) mation on the document image are interpreted and classifLed by lhis system which then p?QduCeS as output a?[ cditablr descl ipticw 0-...

A rule based system for speech verification is proposed as a feedback mechanism for speech recognition. The system provides an explicit way ofselecting verification parameters depending on the item to be recognized and its context. An extended continuous logic is used to combine the evidence from different parameters and an automatic way to convert parameter values to logic evidence values is g...

the concept of intelligently controlling the search process of gravitational search algorithm (gsa) is introduced to develop a novel data mining technique. the proposed method is called fuzzy gsa miner (fgsa-miner). at first a fuzzy controller is designed for adaptively controlling the gravitational coefficient and the number of effective objects, as two important parameters which play major ro...