Search

This paper studies the problem of mining named entity translations by aligning comparable corpora. Current state-of-the-art approaches mine a translation pair by aligning an entity graph in one language to another based on node similarity or propagated similarity of related entities. However, they, building on the assumption of “symmetry”, quickly deteriorate on “weakly” comparable corpora with some asymmetry. In this paper, we pursue two directions for overcoming relation and entity asymmetry respectively. The first approach starts from weakly comparable corpora (for high recall) then ensures precision by selective propagation only to entities of symmetric relations. The second approach starts from parallel corpora (for high precision) then enhances recall by extending the translation matrix based on node similarity and contextual similarity. Our experimental results on English-Chinese corpora show that both approaches are effective and complementary. Our combined approach outperforms the best-performing baseline in terms of F1-score by up to 0.28. [ABSTRACT FROM AUTHOR]

• Delta-T signature-based clustering is proposed for real operation pattern and fault diagnosis. • Operation patterns are analyzed by operational signatures in the normal delta-T area. • Faults are diagnosed by fault signatures in the negative and extreme delta-T areas. • The proposed method is applied to a chilled water cooling system with real operational datasets. Energy consumption patterns are key information for efficiently operating energy systems in buildings or primary energy systems in cities. Nevertheless, operation pattern analysis (e.g., signature analysis), presenting the correlations between system variables, is needed to diagnose inefficient operation or system faults because conventional energy usage patterns have limited insight into the operation patterns inside building energy systems. Thus, a novel operation pattern analysis method is proposed for building energy systems using the clustering-based operational signatures. The novelty of this method is that it is based on delta-T signatures in building energy systems. Because delta-T can be a major representative of operational states, delta-T signature-based clustering can determine both normal and faulty operation. Moreover, delta-T (e.g., water temperature differences between supply and return paths) can be easily measured in a basic sensor deployment in a building automation system. The delta-T signatures can be converted into operational and faulty signatures. These features enable the operation pattern analysis to be applied to building energy systems widely without the need for luxurious sensing systems or building energy management systems. In this study, the proposed method was applied to an operational chilled water cooling system for real multiunit residential buildings. The application shows how the delta-T signature-based clustering method can diagnose operation patterns and various faults. In the target buildings, inefficient operation (accounting for 30.8%) was found in the normal delta-T area, and sensor errors (random, bias, and drift) and abnormal device sequences were captured in the negative or extreme delta-T areas. These patterns and faults could be matched to daily cooling energy usage patterns, thus providing a deeper knowledge of operation. [ABSTRACT FROM AUTHOR]

Abstract: Recently, mesoporous silica was blended with polyimide to develop low dielectric constant (k) materials with improving mechanical and thermal properties of polyimide by utilizing both the nanoporous structure and silica framework. However, even the use of mesoporous silica did not show a significant decrease of k due to the phase segregation in between polyimide and the mesoporous silica materials. In this work, we attempted to prepare polyimide/mesoporous silica hybrid nanocomposites having relatively good phase mixing behavior by utilizing polyimide synthesized from a water soluble poly(amic acid) ammonium salt, which lead to low k up to 2.45. The thermal properties of polyimide were improved by adding mesoporous silicas. For this work, we have fabricated mesoporous silicas through surfactant-templated condensation of tetraethyl orthosilicate (TEOS). Pyromellitic dianhydride (PMDA)-4,4′-oxydianiline (ODA) polyimide was prepared from poly(amic acid) ammonium salt, which had been obtained by incorporating triethylamine (TEA) into PMDA-ODA poly(amic acid) in dimethylacetamide (DMAc), followed by thermal imidization. [Copyright &y& Elsevier]

Abstract: In this work, we prepared and characterized polyamideimide (PAI)/silica hybrids compatibilized with 3-aminopropyltriethoxysilane (APTES). PAI/silica nanohybrid thin films were prepared using an in situ sol–gel process, followed by thermal imidization. We have investigated the microstructures and properties of the PAI/silica hybrids using FT-IR spectroscopy, X-ray diffraction, small-angle X-ray scattering (SAXS), and differential scanning calorimetry (DSC). We also measured their tensile properties, thermal properties, refractive indices, and dielectric constants. In general, the properties of the PAI/silica hybrids were optimized when the silica content was 6wt.%. [Copyright &y& Elsevier]