O operador de agrupamento e as funções de agregação são as principais ferramentas utilizadas para sumarizar dados em um Sistema de Gerenciamento de Base de Dados Relacionais (SGBDR). O operador de agrupamento funciona criando partições nos dados utilizando comparações por identidade, e permite que sejam aplicadas funções de agregação que retornam um único valor representando o grupo como um todo. Entretanto, para dados métricos, agrupamento utilizando identidade tem pouca utilidade. Neste caso, adotar o conceito de similaridade é frequentemente uma abordagem mais promissora. A literatura apresenta alguns operadores que podem agrupar os dados utilizando similaridade. Todos eles utilizam um limiar de valor de distância para atribuir os elementos aos grupos. No entanto, estes operadores não obtêm resultados satisfatórios quando a distribuição dos dados apresenta variações significativas na densidade de objetos em diferentes regiões do espaço. Para alcançar melhores resultados nestas situações, propusemos um novo operador que atribui os grupos utilizando uma eleição envolvendo grupos já atribuídos. Também propusemos generalizações, para os operadores existentes e propostos, para trabalhar com uma quantidade de vizinhos mais próximos e aproximação dos vizinhos mais próximos ao invés de um limiar de distância. Para possibilitar a inclusão destes operadores em SGBDR, propusemos uma extensão à Structured Query Language (SQL) e novas funções de agregação. Implementamos estes operadores em nosso framework em C++ usando a biblioteca Arboretum. Para avaliar os métodos propostos, analisamos tanto qualidade dos resultados quanto tempo de execução, utilizando conjuntos de dados reais e sintéticos. Os operadores propostos alcançaram melhores resultados quanto à qualidade de resultados, e mantiveram os tempos de execução similares. Os operadores que utilizam aproximação aos vizinhos mais próximos produziram resultados de qualidade similar quando comparados aos operadores que utilizando os vizinhos mais próximos, podendo ser executados em menor tempo que estes. / The grouping operator and aggregation functions are the primary tools used to summarize data inside a Relational Database Management Systems (RDBMS). The grouping operator works creating partitions in data using identity comparisons, and allow applying aggregation functions that return a single value that represent the entire group. However, for metric data, grouping by identity is seldom useful. In this case, adopting the concept of the similarity is often a better approach. The literature presents few operators that can group data using similarity. All of them use a distance threshold value to assign the elements in groups. However, these operators do not achieve satisfactory results when the data distribution present a significant variation in the density of objects in different regions of the space. To achieve better results in these situations, we have proposed a novel operator that assign groups using an election involving already assigned groups. We also proposed generalizations to existing and proposed operators to work with an amount of nearest neighbors and approximate neighbors instead of a distance threshold. To support these operators in RDBMS, we propose an extension to Structured Query Language (SQL) and new aggregation functions. Our proposed algorithms can run the proposed and existing operators. We implemented these operators in our framework in C++ using Arboretum library. To evaluate the proposed methods, we assess both results quality and the execution time, using both real and synthetic datasets. The proposed operators achieved better results comparing the quality and maintained similar executing time. The operators that use the approximate nearest neighbors produced similar quality results comparing with the operators that use the exact neighbors and can execute faster than that.

published_or_final_version / abstract / Business / Doctoral / Doctor of Philosophy

Two studies were conducted in order to investigate the role of similarity and influence processes in groups. It was proposed that if group members believe they share one or more salient characteristics (i.e., relevant to the evaluation process) with a target it will (a) fundamentally change the way those group members orient themselves towards that target, and (b) subsequently affect the way those members contribute to the group discussion. Findings suggested that group members who were similar to a target were perceived as having contributed more novel arguments to the group discussion than those dissimilar, and high self-reported levels of communication competence significantly predicted a person's ability to generate arguments about a target. Limitations and future directions are discussed.

Climate variability and landscape characteristics interact to define specific catchment hydrological response. Catchments are considered fundamental landscape units to study the water cycle, since all aspects of the land surface component of the hydrological cycle come together in a defined area, which enables scientific research through mass, momentum and energy budgets. The role of climate-landscape interactions in defining hydrologic partitioning, particularly at the catchment scale, however, is still poorly understood. In this study, a catchment scale process-based hydrologic model (hillslope storage Boussinesq- soil moisture model, hsB-SM) was developed to investigate such interactions. The model was applied to 12 catchments across a climate gradient. Dominant time scales (T.S.) of catchment response and their dimensionless ratios were analyzed with respect to climate and landscape features to identify similarities in catchment response. A limited number of model parameters could be related to observable landscape features. Several T.S. and dimensionless numbers show scaling relationships with respect to the investigated hydrological signatures (runoff coefficient, baseflow index, and slope of the flow duration curve). Some dimensionless numbers vary systematically across the climate gradient, pointing to the possibility that this might be the result of systematic co-variation of climate, vegetation and soil related T.S. Each of 12 behavioral hsB-SM models were subsequently subjected to each of 12 different climate forcings. Mean deviations from Budyko's hypothesis controlling long-term hydrologic partitioning (represented by the evaporation index, E/P, dependence on the aridity index, PET/P) were computed per catchment and per climate. The trend observed per catchment could be explained by the dimensionless ratio of perched aquifer storage-release T.S. and mean storm duration T.S. The trend observed per climate could be explained by an empirical relationship between fraction of rainy days and average daily temperature during those days. Catchments that produce more E/P have developed in climates that produce less E/P, when compared to Budyko's hypothesis. Also, climates that give rise to more E/P are associated with catchments that have vegetation with less efficient water use parameters. These results suggest the possibility of vegetation and soil co-evolution in response to local climate leading to (catchment scale) predictable hydrologic partitioning.

No description available.

This thesis deals with the nature and definition of phonological similarity and shows that, when similarity plays a role in the motivation of phonological processes, it is evaluated over abstract, phonological features and not purely phonetic properties. Empirical evidence for this position is drawn from the domain of consonant harmony. Typological studies of consonant harmony (Hansson 2001, Rose and Walker 2004) have argued that segments which interact in consonant harmony processes must be highly similar to one another. This thesis provides analyses of a range of consonant harmony processes and demonstrates that, in each case, the notion of similarity needed in order to determine participating segments is evaluated over contrastive feature specifications. Contrastive specifications are established according to language-specific feature hierarchies (Jackobson and Halle 1956, Dresher 2003, forthcoming) with some features taking scope over others. Languages analyzed in some detail include Bumo Izon, Kalabari Ijo, Hausa, Dholuo, Anywa, Tzutujil and Aymara. Two definitions of similarity are proposed in order to account for two sets of cases. In one set of consonant harmony processes, interacting segments are similar in the sense that they constitute the natural class of segments contrastively specified in the harmonic feature. In another set of cases, participating segments must be similar according to the following definition; they must differ in only a single marked and contrastive feature specification.

圖像相似性度量準則是一個傳統的研究領域。大量經典的圖像處理技術被用來為各種類型的圖像設計相似性度量準則，這些圖像包括了線條圖，灰度圖，彩圖以及高動態範圍圖像。儘管已有的度量準則在指定的條件下可以實現優良的圖像相似度比較，這些度量準則極少系統地考慮或檢驗自身與人類視覺感知之間的一致性。而與人類知覺的一致性是由大量實際應用提出的共同需求。隨著三維立體設備的廣泛應用，圖像的相似性已經不只是傳統的可視差別，更包括了人眼利用三維立體設備同時觀看兩張不同的圖片時的視覺可接受度。 / 非嚴謹對準形狀相似性度量準則（AISS）可以比較兩幅具有固定尺寸的線條圖的形狀相似度。對於該度量準則，兩幅待比較圖像的形狀不要求完全對齊，同時，又會考慮到圖像的形變，例如位置，方向和縮放上的變化。 / 雙目觀看舒適度預測器（BVCP）是另一個度量準則。當人的雙眼同時觀看兩幅不同的圖像時，該準則可用以預測視覺的舒適度。根據著名的双眼單视理論，人的視覺可以將兩幅具有細節、對比度以及亮度差別的圖像合成一幅圖像，只要這些差別在限定的程度之內。在計算機圖形學領域，BVCP 首次嘗試去預測雙目的圖像差別會否引起觀看的不舒適。 / 在本論文中，實用的應用程序也被提出用以衡量AISS 和BVCP。AISS 被用在了一個名為“基於結構的ASCII 藝術的應用程序中，該應用程序可以利用ASCII 字符的形狀近似地表現參考圖像的線條結構信息。而BVCP 則被用在一個創新地應用框架中，該框架可以從單幅高動態範圍圖像中生成一組（兩幅）低動態範圍圖像。當這一組低動態範圍圖像組被人的雙眼同時觀看時，可以比傳統的單幅低動態範圍圖像保留更多的人類可感知視覺信息。可信的結果和使用者研究也用來證明SSIM 和BVCP 的有效性以及與人類知覺的一致性。 / Image similarity metric is a traditional research field. Classical image processing techniques are used to design similarity metrics for all kinds of images, such as line drawings, gray or color image and even high-dynamic range (HDR) images. While existing metrics perform well for the tasks of comparing images in specified situations, few of them have systematically considered or examined the consistency with human perception required by practical applications. With the blooming of stereo devices, the similarity to be measured is not only the traditional visual difference between two images, but also the visual acceptance of two images when they are viewed simultaneously with 3D devices. This thesis presents two image similarity metrics motivated by perceptual principles, also with applications to demonstrate their novelty and practical values. / Alignment-Insensitive Shape Similarity Metric (AISS) measures shape similarity of line drawings. This metric can tolerate misalignment between two shapes and, simultaneously, accounts for the differences in transformation such as, position, orientation and scaling. / Binocular Viewing Comfort Predictor (BVCP) is another metric proposed to measure visual discomfort when human's two eyes view two different images simultaneously. According to a human vision phenomenon - binocular single vision, human vision is able tofuse two images with differences in detail, contrast and luminance, up to a certain limit. BVCP makes a first attempt in computer graphics to predict such visual comfort limit. / Applications are also proposed to evaluate AISS and BVCP. AISS is utilized in an application of Structure-based ASCII Art, which approximates line structure of the reference image content with the shapes of ASCII characters. BVCP is utilized in a novel framework - Binocular Tone Mapping which generates a binocular low-dynamic range (LDR) image pair from one HDR image. Such binocular LDR pair can be viewed with stereo devices and can preserve more human-perceivable visual content than traditional one single LDR image. Convincing results and user studies are also shown to demonstrate that both AISS and BVCP are consistent with human perception and effective in practical usage. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Zhang, Linling. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 122-132). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese. / Abstract --- p.i / Acknowledgement --- p.v / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Alignment-Insensitive Shape Similarity Metric --- p.8 / Chapter 2.1 --- Related Work --- p.10 / Chapter 2.2 --- Design of AISS --- p.13 / Chapter 2.2.1 --- Misalignment Tolerance --- p.14 / Chapter 2.2.2 --- Transformation Awareness --- p.16 / Chapter 2.2.3 --- Parameter Setting --- p.17 / Chapter 2.3 --- Results and Discussion --- p.18 / Chapter 2.4 --- Discussion --- p.20 / Chapter 3 --- Application for AISS: Structure-based ASCII Art --- p.21 / Chapter 3.1 --- Overview --- p.24 / Chapter 3.2 --- Optimization --- p.28 / Chapter 3.3 --- User Study and Discussion --- p.35 / Chapter 3.3.1 --- Metrics Comparison --- p.35 / Chapter 3.3.2 --- Comparison to Existing Work --- p.38 / Chapter 3.3.3 --- User Study --- p.40 / Chapter 3.4 --- Summary --- p.44 / Chapter 4 --- Binocular Viewing Comfort Predictor --- p.48 / Chapter 4.1 --- Background --- p.51 / Chapter 4.2 --- Design of BVCP --- p.54 / Chapter 4.2.1 --- Fusional Area --- p.55 / Chapter 4.2.2 --- Contour Fusion --- p.58 / Chapter 4.2.3 --- Contour and Regional Contrasts --- p.68 / Chapter 4.2.4 --- Failure of Rivalry --- p.70 / Chapter 4.2.5 --- The Overall Fusion Predictor --- p.74 / Chapter 4.3 --- User Study --- p.77 / Chapter 4.4 --- Discussion and Limitations --- p.84 / Chapter 5 --- Application for BVCP: Binocular Tone Mapping --- p.86 / Chapter 5.1 --- Framework --- p.90 / Chapter 5.1.1 --- Overview --- p.90 / Chapter 5.1.2 --- Optimization --- p.93 / Chapter 5.2 --- Results and Discussion --- p.96 / Chapter 5.2.1 --- Results --- p.96 / Chapter 5.2.2 --- User Study --- p.103 / Chapter 5.2.3 --- Incorporating Stereopsis --- p.106 / Chapter 5.2.4 --- Limitations --- p.109 / Chapter 5.3 --- Summary --- p.112 / Chapter 6 --- Conclusion --- p.113 / Chapter A --- User Study for ASCII art --- p.117 / Bibliography --- p.122

Thesis (Ph. D.)--University of Hong Kong, 2007. / Title proper from title frame. Also available in printed format.

Group comparison per se is a fundamental task in many scientific endeavours but is also the basis of any classifier. Comparing groups of sequence data is a relevant task. To contrast sequence groups, we define Emerging Sequences (ESs) as subsequences that are frequent in sequences of one group and less frequent in another, and thus distinguishing sequences of different classes. There are two challenges to distinguish sequence classes by ESs: the extraction of ESs is not trivially efficient and only exact matches of sequences are considered. In our work we address those problems by a suffix tree-based framework and a sliding window matching mechanism. A classification model based on ESs is also proposed. Evaluating against several other learning algorithms, the experiments on two datasets show that our similar ESs-based classification model outperforms the baseline approaches. With the ESs' high discriminative power, our proposed model achieves satisfactory F-measures on classifying sequences.

Thesis (Ph. D.)--University of Hong Kong, 2007. / Also available online.