Learning to match ontologies on the Semantic Web

On the Semantic Web, data will inevitably come from many different ontologies, and information processing across ontologies is not possible without knowing the semantic mappings between them. Manually finding such mappings is tedious, error-prone, and clearly not possible on the Web scale. Hence the development of tools to assist in the ontology mapping process is crucial to the success of the Semantic Web. We describe GLUE, a system that employs machine learning techniques to find such mappings. Given two ontologies, for each concept in one ontology GLUE finds the most similar concept in the other ontology. We give well-founded probabilistic definitions to several practical similarity measures and show that GLUE can work with all of them. Another key feature of GLUE is that it uses multiple learning strategies, each of which exploits well a different type of information either in the data instances or in the taxonomic structure of the ontologies. To further improve matching accuracy, we extend GLUE to incorporate commonsense knowledge and domain constraints into the matching process. Our approach is thus distinguished in that it works with a variety of well-defined similarity notions and that it efficiently incorporates multiple types of knowledge. We describe a set of experiments on several real-world domains and show that GLUE proposes highly accurate semantic mappings. Finally, we extend GLUE to find complex mappings between ontologies and describe experiments that show the promise of the approach.Received: 16 December 2002, Accepted: 16 April 2003, Published online: 17 September 2003Edited by: Edited by B.V. Atluri, A. Joshi, and Y. Yesha     

Behavior of Compacted Soil-Fly Ash-Carbide Lime Mixtures

Unconfined compression tests, Brazilian tensile tests, and saturated drained triaxial compression tests with local strain measurement were carried out to evaluate the stress-strain behavior of a sandy soil improved through the addition of carbide lime and fly ash. The effects of initial and pozzolanic reactions were investigated. The addition of carbide lime to the soil-fly ash mixture caused short-term changes due to initial reactions, inducing increases in the friction angle, in the cohesive intercept, and in the average modulus. Such improvement might be of fundamental importance to allow site workability and speeding construction purposes. In addition, under the effect of initial reactions, the maximum triaxial stiffness occurred for specimens molded on the dry side of the optimum moisture content, while the maximum strength occurred at the optimum moisture content. After 28 days, pozzolanic reactions magnified brittleness and further increased triaxial peak strength and stiffness; the maximum triaxial strength and stiffness occurred on the dry side of the optimum moisture content.     

Chemical and morphological features of nanofilled composite resin: influence of finishing and polishing procedures and fluoride solutions

This study evaluated the influence of finishing and polishing procedures and different fluoride solutions on superficial morphology and chemistry of the nanofilled composite resin Supreme XT (3M) through the EDX analysis and SEM evaluation. Circular specimens (n = 30) of 10 mm diameter and 2 mm thickness were prepared, with half of the sample assays finished and polished with Super‐Snap® sandpaper. The experimental groups were divided according to the presence or absence of finishing and polishing and solutions (artificial saliva, 0.05% of manipulated sodium fluoride solution, Fluordent Reach, Oral B, Fluorgard). Specimens were immersed in each respective solution for 1 min per day, during 60 days and stored in artificial saliva at 37 ± 1°C between immersion periods. Topography and chemical analysis was qualitative. It was observed that specimens submitted to finishing and polishing procedures had lower superficial degradation. Fluoride solutions promoted superficial alterations on specimens, being the highest degradation obtained with Fluordent Reach. It can be concluded that finishing and polishing procedures and the immersion media influence the superficial morphology of composite resin tested; the Fluordent Reach was the fluoride solution that most affected the material's surface. Microsc. Res. Tech. 2011., © 2011 Wiley Periodicals, Inc.     

Enhancing the recovery and concentration of polyphenols from camu-camu (Myrciaria dubia H.B.K. McVaugh) by aqueous two-phase flotation and scale-up process

In this study, polyphenols obtained from camu-camu pulp residue were recovered and concentrated from aqueous two-phase flotation (ATPF) by evaluating different operating conditions, such as solvent type, pH, ammonium sulfate concentration, flotation time, air flow, and volumetric ratio. Under optimal conditions, the polyethylene glycol 400/ammonium sulfate ATPF system showed an efficiency of 81.02% and a concentration coefficient two times higher than traditional aqueous two-phase extraction. Experiments were also carried out on a semi-pilot scale, using an apparatus with a 2.5-L working volume. The results showed that the semi-pilot system was comparable to the smaller scale system.     

Artificial Neural Networks Associated to Calorimetric Measurements Used as a Method to Predict Polymer Composition of High Solid Content Emulsion Copolymerizations

Summary: Inspired by biological systems, artificial neural networks (ANN) have demonstrated to be powerful tools to model non‐linear systems, such as high solid content latexes produced by emulsion polymerization which have a great importance in the polymeric industry, essentially for environmental reasons, since they usually have water as a continuous phase. The quality of the produced polymer is closely related to the structure of the polymeric chain. In order to propose technical and economically feasible alternatives to control a polymeric structure, this work is aimed to develop a new methodology based on ANN associated with calorimetry to predict the polymeric structure. The designed ANN presented excellent results when tested with process condition variations (such as temperature and reaction time) as well as when they were submitted to test concerning the variation on the proportion of monomers in the latex formulation. Hence, it was possible to conclude that ANN, associated to calorimetry, lead to an efficient method to predict the polymer composition in emulsion copolymerizations.

     

Markov logic networks

We propose a simple approach to combining first-order logic and probabilistic graphical models in a single representation. A Markov logic network (MLN) is a first-order knowledge base with a weight attached to each formula (or clause). Together with a set of constants representing objects in the domain, it specifies a ground Markov network containing one feature for each possible grounding of a first-order formula in the KB, with the corresponding weight. Inference in MLNs is performed by MCMC over the minimal subset of the ground network required for answering the query. Weights are efficiently learned from relational databases by iteratively optimizing a pseudo-likelihood measure. Optionally, additional clauses are learned using inductive logic programming techniques. Experiments with a real-world database and knowledge base in a university domain illustrate the promise of this approach. Editors: Hendrik Blockeel, David Jensen and Stefan Kramer An erratum to this article is available at .     

The Role of Occam's Razor in Knowledge Discovery

Many KDD systems incorporate an implicit or explicit preference for simpler models, but this use of Occam's razor has been strongly criticized by several authors (e.g., Schaffer, 1993; Webb, 1996). This controversy arises partly because Occam's razor has been interpreted in two quite different ways. The first interpretation (simplicity is a goal in itself) is essentially correct, but is at heart a preference for more comprehensible models. The second interpretation (simplicity leads to greater accuracy) is much more problematic. A critical review of the theoretical arguments for and against it shows that it is unfounded as a universal principle, and demonstrably false. A review of empirical evidence shows that it also fails as a practical heuristic. This article argues that its continued use in KDD risks causing significant opportunities to be missed, and should therefore be restricted to the comparatively few applications where it is appropriate. The article proposes and reviews the use of domain constraints as an alternative for avoiding overfitting, and examines possible methods for handling the accuracy–comprehensibility trade-off.     

Piecewise polynomial dielectric function model and its application for the retrieval of optical functions

A new expression of dielectric function model based on piecewise polynomials is introduced. Its association with spline and more recent shape preserving interpolation algorithms allows easy reproduction of every kind of experimental spectra and thus retrieval of all the linear optical functions of a material. Based on a pure mathematical framework, the expression of the model is always applicable and does not necessitate any knowledge of the microscopic mechanisms of absorption responsible for the optical response. The potential of piecewise polynomial dielectric functions is shown through synthetic examples and the analysis of experimental spectra.     

Preliminary virulence genotyping and phylogeny of Escherichia coli from the gut of pigs at slaughtering stage in Brazil

In the current study we screened Escherichia coli from intestine of pigs slaughtered in Mato Grosso, Brazil, for virulence-markers related to human disease. Furthermore, we employed for the first time a phylogenetic assay to explore the association between phylogeny and virulence genotype in E. coli from finished swine. A low prevalence (7.8%) of E. coli harbouring virulence genes was observed. Among the positive isolates, 3.3% could be classified as atypical EPEC, 2.2% as STEC and 2.2% as CDT harbouring E. coli. Virulence genes were not found to co-occur in a strain. Phylogenetic determination of isolates revealed a low prevalence of E. coli lineages related to disease. Therefore, preliminary sampling of 74 pigs indicated that slaughter swine may not be major reservoirs of E. coli capable of causing human disease. In light of the significant association between phylogeny and virulence genotype, we also underscored the phylogenetic grouping of strains as a valuable tool for E. coli surveillance programmes in slaughterhouses.