Structure-sensitive semantic matching for aggregate question answering over knowledge base

Aggregate question answering essentially returns answers for given questions by obtaining query graphs with unique dependencies between values and corresponding objects. Word order dependency, as the key to uniquely identify dependency of the query graph, reflects the dependencies between the words in the question. However, due to the semantic gap caused by the expression difference between questions encoded with word vectors and query graphs represented with logical formal elements, it is not trivial to match the correct query graph for the question. Most existing approaches design more expressive query graphs for complex questions and rank them just by directly calculating their similarities, ignoring the semantic gap between them. In this paper, we propose a novel Structure-sensitive Semantic Matching(SSM) approach that learns aligned representations of dependencies in questions and query graphs to eliminate their gap. First, we propose a cross-structure matching module to bridge the gap between two modalities(i.e., textual question and query graph). Then, we propose an entropy-based gated AQG filter to remove the structural noise caused by the uncertainty of dependencies. Finally, we present a two-channel query graph representation that fuses the semantics of abstract structure and grounding content of the query graph explicitly. Experimental results show that SSM could learn aligned representations of questions and query graphs to eliminate the gaps between their dependencies, and improves up to 12% (F1 score) on aggregation questions of two benchmark datasets.     

Manufacturing catalyst-coated membranes by ultrasonic spray deposition for PEMFC: Identification of key parameters and their impact on PEMFC performance

This study deals with the manufacturing of catalyst-coated membranes (CCMs) for newcomers in the field of coating. Although there are many studies on electrode ink composition for improving the performance of proton-exchange membrane fuel cells (PEMFCs), there are few papers dealing with electrode coating itself. Usually, it is a know-how that often remains secret and constitutes the added value of scientific teams or the business of industrialists. In this paper, we identify and clarify the role of key parameters to improve coating quality and also to correlate coating quality with fuel cell performance via polarization curves and electrochemical active surface area measurements. We found that the coating configurations can affect the performance of lab-made CCMs in PEMFCs. After the repeatability of the performance obtained by our coating method has been proved, we show that: (i) edge effects, due to mask shadowing - cannot be neglected when the active surface area is low, (ii) a heterogeneous thickness electrode produces performance lower than a homogeneous thickness electrode, and (iii) the origin and storage of platinum on carbon powders are a very important source of variability in the obtained results.     

Effect of 100 KeV Ar+ ion beam irradiation on ZnO thin films – Influence of morphology vis-a-vis electrode/electrolyte interface and its impact on photoelectrochemical water splitting

The present study attempts quantitative determination of changes in the morphological surface features viz. fractal dimension, lower and upper cut off length scale through Power Spectral Density analysis prior to and after irradiation of 100 KeV Ar⁺ ion beam at incidence angles of 0°, 40° and 60° on ZnO thin films. All the unirradiated and irradiated samples are subjected to photoelectrochemical characterization and a correlation between photoelectrochemical performance and morphological parameters is established. Sample irradiated at 40° angle at the fluence of 5 × 10¹⁶ ions/cm² is found to possess maximum fractal dimension of 2.72, lower and upper cut off length scale of 3.16 nm and 63.00 nm respectively. This sample exhibits maximum photocurrent density of 3.19 mA/cm² and applied bias photon-to-current efficiency of 1.12% at 1.23 V/RHE. Hydrogen gas collected for duration of 1 h for the same sample was ~4.83 mLcm^?2.     

Reliability and scale effect in toughness of concrete structures

In the present work, the distribution of the random toughness characteristics (i.e. critical energy release rate, G_1c) has been evaluated on the basis of experimental observations. Fracture test results from three groups of geometrically similar concrete specimens of size (width×total depth×thickness), 420×420×50–1680×1680×200 mm³, made with different maximum aggregate size of 9.5, 19, 38, and 76 mm were analyzed using a recently proposed distribution of extremes. In applications of probability, it is important to use an appropriate distribution type and adequate techniques for estimating the parameters of distribution. In this study, a new type distribution of minima is employed for probability computations. It was noticed that the entropy of distribution increases with the crack length, i.e. the uncertainty of toughness, G_1c, value increases with crack length. A non-linear reduction of the maximum allowable splitting force with the defect size, a, was noticed. For large specimens, the maximum allowable splitting load is more sensitive to the required reliability level than that for small specimens. Reliability increases with aggregate size when all other conditions were constant.     

Compile-time copy elimination

Single-assignment and functional languages have value semantics that do not permit side-effects. This lack of side-effects makes automatic detection of parallelism and optimization for data locality in programs much easier. However, the same property poses a challenge in implementing these languages efficiently. This paper describes an optimizing compiler system that solves the key problem of aggregate copy elimination. The methods developed rely exclusively on compile-time algorithms, including interprocedural analysis, that are applied to an intermediate data flow representation. By dividing the problem into update-in-place and build-in-place analysis, a small set of relatively simple techniques—edge substitution, graph pattern matching, substructure sharing and substructure targeting—was found to be very powerful. If combined properly and implemented carefully, the algorithms eliminate unnecessary copy operations to a very high degree. No run-time overhead is imposed on the compiled programs.     

人支气管上皮靶细胞形态学参数的测量

支气管上皮靶细胞形态学参数对计算受氡子体照射的支气管的剂量是很重要的。本文介绍利用奥尔新兰-PAS 染色的外科切除人肺组织切片,测量基底细胞粘液细胞和的胞核深度与数目。结果表明,叶、段和小支气管的粘液细胞核深度分别为相应节段基底细胞的61%、69%和73%;此三节段粘液细胞的数目分别为基底细胞的53%、29%和28%,平均为37%。     

Incremental maintenance of aggregate and outerjoin expressions

Views stored in a data warehouse need to be kept current. As recomputing the views is very expensive, incremental maintenance algorithms are required. Over recent years, several incremental maintenance algorithms have been proposed. None of the proposed algorithms handle the general case of relational expressions involving aggregate and outerjoin operators efficiently.     

生物非光滑耐磨表面与非光滑耐磨技术在工程中的应用

指出生物非光滑耐磨表面现象的普遍性，对贝壳、牙齿等几种天然生物耐磨表面几何形态特征进行了分析，并介绍了加工耐磨表面的几种技术方法及非光滑耐磨技术在工程中的应用。同时，对非光滑耐磨理论与技术方法中存在的问题及发展趋势进行了分析与展望。     

视频图像活动轮廓目标检测跟踪研究

研究视频图像目标跟踪定位精确度问题。由于在图像中通常会发生缩放,造成图像目标模糊不清。传统的目标跟踪算法该类算法仅以目标发生平移运动为假设前提,图像质量差。为解决上述问题,提出了一种活动轮廓目标跟踪定位检测算法。首先选择合适的滑窗,采用减背景法来确定视频对象的运动区域,采用卡尔曼形态滤波来消除残余的噪声,然后针对目标在活动轮廓局部内具有较高灰度值的特征,通过自适应阈值来判别滑窗中心位置是否存在目标。当滑窗遍历整幅图像后,就可以得到目标的定位结果。仿真结果表明,改进算法不仅能够消除差分图像中的显露背景,从而得到运动视频对象精确的轮廓,并且可进行多目标的分割与跟踪,具有一定的实际应用价值。     

采用形态神经网络背景自适应预测的图像弱小目标检测*

提出一种有效的背景杂波预测形态神经网络模型,用于检测图像数据中的弱小目标.目标被假设为只有很小的空域扩展度,而且淹没于强背景杂波干扰中.通过形态神经网络,杂波背景被准确地估计并从输入数据中去除,只剩下残留噪声和目标信号.采用扩展输入层数据的办法修正了传统的形态开、闭运算三层前馈BP网络模型.为了跟踪包含不同子结构的复杂背景,原始图像被划分为多个子块,并在相应的子块中选择训练样本对结构元进行优化.对真实图像数据的计算机仿真表明该算法在性能上优于其他传统算法.