Relational similarity-based model of data part 1: foundations and query systems

We present a general rank-aware model of data which supports handling of similarity in relational databases. The model is based on the assumption that in many cases it is desirable to replace equalities on values in data tables by similarity relations expressing degrees to which the values are similar. In this context, we study various phenomena which emerge in the model, including similarity-based queries and similarity-based data dependencies. Central notion in our model is that of a ranked data table over domains with similarities which is our counterpart to the notion of relation on relation scheme from the classical relational model. Compared to other approaches which cover related problems, we do not propose a similarity-based or ranking module on top of the classical relational model. Instead, we generalize the very core of the model by replacing the classical, two-valued logic upon which the classical model is built by a more general logic involving a scale of truth degrees that, in addition to the classical truth degrees 0 and 1, contains intermediate truth degrees. While the classical truth degrees 0 and 1 represent nonequality and equality of values, and subsequently mismatch and match of queries, the intermediate truth degrees in the new model represent similarity of values and partial match of queries. Moreover, the truth functions of many-valued logical connectives in the new model serve to aggregate degrees of similarity. The presented approach is conceptually clean, logically sound, and retains most properties of the classical model while enabling us to employ new types of queries and data dependencies. Most importantly, similarity is not handled in an ad hoc way or by putting a “similarity module” atop the classical model in our approach. Rather, it is consistently viewed as a notion that generalizes and replaces equality in the very core of the relational model. We present fundamentals of the formal model and two equivalent query systems which are analogues of the classical relational algebra and domain relational calculus with range declarations. In the sequel to this paper, we deal with similarity-based dependencies.     

Structure and mechanical properties of plasma sprayed coatings of titania and alumina

The paper concerns the use of traditional and depth-sensing indentation (DSI) for investigation of deposits produced from powders based on conventional and nano-sized particles by plasma spray technology.

Plasma sprayed coatings of titania and alumina were studied. Polished cross-section of each coating was prepared and matrices of nano-indents with Berkovich tip were applied onto both materials to explore local elastic behavior. Applied load was in the range of mN to create indents with the same size scale as the thickness of splats—the main building units of the coating. The hardness value as well as the load/unload curve for each indent was stored. Titania coating was sprayed from a novel type of nanoscale-size powder agglomerated to particles useful for plasma spraying, whereas fused and crushed conventional powder was utilized for alumina spraying and for titania coating used as reference. The effect of annealing on elastic properties of titania was studied as well. The values of elastic parameters as well as the character of the coating inhomogeneity seem to reflect: (i) the composition of material and the fabrication technique and (ii) microstructural differences between coatings that are partly inherited from the feedstock powders. The results of DSI tests are discussed also in comparison with common technique used for the investigation of plasma coatings hardness—Vickers microhardness measurement.     

Impact of Boolean factorization as preprocessing methods for classification of Boolean data

We explore a utilization of Boolean matrix factorization for data preprocessing in classification of Boolean data. In our previous work, we demonstrated that preprocessing that consists in replacing the original Boolean attributes by factors, i.e. new Boolean attributes obtained from the original ones by Boolean matrix factorization, can improve classification quality. The aim of this paper is to explore the question of how the various Boolean factorization methods that were proposed in the literature impact the quality of classification. In particular, we compare five factorization methods, present experimental results, and outline issues for future research.     

Mechanical Properties and Microstructural Characterization of Amorphous SiCxNy Thin Films After Annealing Beyond 1100°C

The effect of thermal annealing on structure and mechanical properties of amorphous SiC_xN_y (y ≥ 0) thin films was investigated up to 1500°C in air and Ar. The SiC_xN_y films (2.2–3.4 μm) were deposited by reactive DC magnetron sputtering on Si, Al₂O₃ and α‐SiC substrates without intentional heating and at 600°C. The SiC target with small excess of carbon was sputtered at various N₂/Ar gas flow ratios (0–0.48). The nitrogen content in the films changes in the range 0–43 at.%. Hardness and elastic modulus (nanoindentation), change in film thickness, film composition, and structure (Raman spectroscopy, XRD) were investigated in dependence on annealing temperature and nitrogen content. All SiC_xN_y films preserve their amorphous structure up to 1500°C. The hardness of all as‐deposited and both air‐ and Ar‐annealed SiC_xN_y films decreases with growth of nitrogen content. The annealing in Ar at temperatures of 1100°C–1300°C results in noticeable hardness growth despite the ordering of graphite‐like structure in carbon clusters in nitrogen free films. Unlike the SiC, this graphitization leads to hardness saturation of SiCN films starting above 900°C, especially for films with higher nitrogen content (deposited at higher N₂/Ar). This indicates the practical hardness limit achievable by thermal treatment for SiC_xN_y films deposited on unheated substrates. The ordering in carbon phase is facilitated by the presence of nitrogen in the films and its extent is controlled by the N/C atomic ratio. The suppression of graphitization was observed for N/C ranging between 0.5–0.7. Films deposited at 600°C show higher hardness and oxidation resistance after annealing in comparison with those deposited on unheated substrates. Hardness reaches 40 GPa for SiC and ~28 GPa for SiC_xN_y (35 at.% of nitrogen). Such a high hardness of SiC film stems from its partial crystallization. Annealing of SiC_xN_y film (35 at.% of N) in Ar at 1400°C is accompanied by formation of numerous hillocks (indicating heterogeneous structure of amorphous films) and redistribution of film material.     

Efficient and environmentally friendly energy systems for microregions

Maximum possible utilization of renewable sources as well as minimization of waste production and preference of recycling is definitely a positive trend. Global economic crisis, however, has brought about several aspects that have been neglected before. It is difficult to find a market for much of the sorted PET bottles and paper; thus sorting itself and recycling unfortunately lose their meaning. This gives rise to the question as to whether current procedures and technologies are appropriate or not. This article tries to offer a new approach which lies in maximizing waste processing (sorting, recycling, incineration, etc.) at the site of its production, along with an emphasis on maximizing efficiency of waste treatment by minimization of transport distances, labour costs and use of synergic effects of several different technologies in an open central controlled system called “MICROREGION”.     

三视校正的理论及鲁棒性算法

主要讨论两方面的工作.首先,对三视校正问题进行理论分析,给出了校正后图像的基本矩阵与其约束条件之间的关系,讨论了三视校正过程中的6个自由参数的几何含义.这些结果为处理校正过程中带来的图像射影畸变提供了理论根据.其次,在RANSAC(random sampling consensus)框架下,提出了一种鲁棒的三视校正算法.与传统的校正算法不同,该算法不再只依赖于基本矩阵,而是直接利用了原始匹配点的信息.这种基于点的方法有两个优点:一方面,由于噪声的干扰,基本矩阵往往估计得不够准确;另一方面,由于基本矩阵的评价准则与校正结果的评价准则不同,即使从好的基本矩阵出发,也未必能获得好的校正结果.大量的模拟和真实图像实验表明,该算法具有很强的抗噪声及抗错误匹配的能力,能够获得令人满意的校正效果.