Joaquín Cañadas

Approximate Algorithms for Distance-Based Queries in High-Dimensional Data Spaces Using R-Trees

In modern database applications the similarity or dissimilarity of complex objects is examined by performing distance-based queries (DBQs) on data of high dimensionality. The R-tree and its variations are commonly cited multidimensional access methods that can be used for answering such queries. Although, the related algorithms work well for low-dimensional data spaces, their performance degrades as the number of dimensions increases (dimensionality curse). In order to obtain acceptable response time in high-dimensional data spaces, algorithms that obtain approximate solutions can be used. Three approximation techniques (?-allowance, N-consider and M-consider) and the respective recursive branch-and-bound algorithms for DBQs are presented and studied in this paper. We investigate the performance of these algorithms for the most representative DBQs (the K-nearest neighbors query and the K-closest pairs query) in high-dimensional data spaces, where the point data sets are indexed by tree-like structures belonging to the R-tree family: R*- trees and X-trees. The searching strategy is tuned according to several parameters, in order to examine the trade-off between cost (I/O activity and response time) and accuracy of the result. The outcome of the experimental evaluation is the derivation of the outperforming DBQ approximate algorithm for large high-dimensional point data sets.

Defining the semantics of rule-based Web applications through model-driven development

Defining the semantics of rule-based Web applications through model-driven development Rule languages and inference engines incorporate reasoning capabilities to Web information systems. This paper presents an approach for the specification and development of Web applications performing the usual functionalities of data management and incorporating a rule engine for reasoning capabilities. The proposed approach is based on the definition of a high-level representation of the semantics of rule-based applications through a formalism for conceptual modeling combining lightweight ontologies and production rules. These models are used as the source for a model-driven method that applies several transformations to conceptual models generating the rule-based Web application code in an automatic process. As a result, the rule-based Web application embeds a rule engine suitable for deducing information by applying an inference process. The structure of the information managed by the Web application is based on ontology classes, whereas the logical expressions applied in reasoning are obtained from production rules of the model. A rule-based Web application has been developed and evaluated using a supporting tool that implements the ideas presented in this paper.

InSCo-Gen: A MDD Tool for Web Rule-Based Applications

Rules and ontologies are widely used in software development since they provide semantic web applications with meaning and reasoning features. This demonstration paper presents InSCo-Gen , a Model-Driven Development (MDD) tool for Web rule-based applications, which constructs a functional Web architecture integrating a rule engine for reasoning tasks. Development process is based on conceptual models composed of ontologies and production rules. These models are the source for the MDD process, which automatically generates implementation of the Web application.

A proposal for Projekt Management Using CommonKADS

This work describes a proposal for software lifecycle and project management when developing knowledge based systems using CommonKADS methodology. To build this proposal, an analysis of the similarities between software engineering and knowledge engineering was made. As a result, the anchor point milestones and the invariant definitions from the spiral model lifecycle for software development have been translated to CommonKADS methodology. This proposal is being applied to a project in order to develop a phytosanitary advisor system used by farmers in greenhouse crops.

Processing distance-based queries in multidimensional data spaces using R-trees

In modern database applications the similarity, or dissimilarity of data objects is examined by performing distance-based queries (DBQs) on multidimensional data. The R-tree and its variations are commonly cited multidimensional access methods. In this paper, we investigate the performance of the most representative distance-based queries in multidimensional data spaces, where the point datasets are indexed by tree-like structures belonging to the R-tree family. In order to perform the K-nearest neighbor query (K-NNQ) and the K-closest pair query (K-CPQ), non-incremental recursive branch-and-bound algorithms are employed. The K-CPQ is shown to be a very expensive query for datasets of high cardinalities that becomes even more costly as the dimensionality increases. We also give ɛ-approximate versions of DBQ algorithms that can be performed faster than the exact ones, at the expense of introducing a distance relative error of the result. Experimentation with synthetic multidimensional point datasets, following Uniform and Gaussian distributions, reveals that the best index structure for K-NNQ is the X-tree. However, for K-CPQ, th e R*-tree outperforms th e X-tree in respect to the response time and the number of disk accesses, when an LRU buffer is used. Moreover, the application of the ɛ-approximate technique on the recursive K-CPQ algorithm leads to acceptable approximations of the result quickly, although the tradeoff between cost and accuracy cannot be easily controlled by the users.

Development of a web tool for action threshold evaluation in table grape pest management

Pest control is a major issue in agricultural management due to crop yield losses caused by pests. In this context, integrated pest management aims to suppress pest populations below an action threshold to minimize their impact. This paper presents the development of a web tool based on the Spanish regulations for the integrated pest management of table grapes; this provides decision support for evaluating when a particular pest action threshold has been crossed thus affecting table grape crops. The tool was built using a model-driven software development approach that enables software system generation from the problem’s knowledge model. The design of the knowledge bases which contain the system’s decision rules is also described. It is divided into knowledge bases that contain general knowledge related to the table grape crop as well as several specific knowledge bases (one per pest) containing the reasoning model that deduces the risk associated with a particular pest. The software has been designed by applying the model-driven development method thus making the system flexible, easy to evolve and adaptable whenever a new pest has to be incorporated into the software.

RuleModeler: a rich internet application for knowledge modeling

Traditionally, the use of standard Web-based architecture for knowledge modeling and management suffers from important weaknesses. This work describes the use of current technology of Rich Internet Applications (RIAs) in tools for knowledge modeling and management. As a proof of the concept we have implemented RuleModeler, a RIA tool prototype for knowledge modeling where knowledge models composed of ontologies and production rules can be specified in a distributed way. The tool has been used to create and maintain the knowledge model for a pest control decision-support system in agriculture.