Exploring automatic feature selection for transition-based dependency parsing

En este artículo se investigan técnicas automáticas para encontrar un modelo óptimo de características en el caso de un analizador de dependencias basado en transiciones. Mostramos un estudio comparativo entre algoritmos de búsqueda, sistemas de validación y reglas de decisión demostrando al mismo tiempo que usando nuestros métodos es posible conseguir modelos complejos que proporcionan mejores resultados que los modelos que siguen configuraciones por defecto.

Viability study of automobile shredder residue as fuel

Car Fluff samples collected from a shredding plant in Italy were classified based on particle size, and three different size fractions were obtained in this way. A comparison between these size fractions and the original light fluff was made from two different points of view: (i) the properties of each size fraction as a fuel were evaluated and (ii) the pollutants evolved when each size fraction was subjected to combustion were studied. The aim was to establish which size fraction would be the most suitable for the purposes of energy recovery. The light fluff analyzed contained up to 50 wt.% fines (particle size < 20 mm). However, its low calorific value and high emissions of polychlorinated dioxins and furans (PCDD/Fs), generated during combustion, make the fines fraction inappropriate for energy recovery, and therefore, landfilling would be the best option. The 50–100 mm fraction exhibited a high calorific value and low PCDD/F emissions were generated when the sample was combusted, making it the most suitable fraction for use as refuse-derived fuel (RDF). Results obtained suggest that removing fines from the original ASR sample would lead to a material product that is more suitable for use as RDF.