Cluster-based information retrieval systems often use a similarity measure to compute the association among text documents. In this thesis, we focus on a class of similarity measures named Query-Sensitive Similarity (QSS) measures. Recent studies have shown QSS measures to positively influence the outcome of a clustering procedure. These studies have used QSS measures in conjunction with the ltc term-weighting scheme. Several term-weighting schemes have superseded the ltc term-weighing scheme and demonstrated better retrieval performance relative to the latter. We test whether introducing one of these schemes, INQUERY, will offer any benefit over the ltc scheme when used in the context of QSS measures. The testing procedure uses the Nearest Neighbor (NN) test to quantify the clustering effectiveness of QSS measures and the corresponding term-weighting scheme. The NN tests are applied on certain standard test document collections and the results are tested for statistical significance. On analyzing results of the NN test relative to those obtained for the ltc scheme, we find several instances where the INQUERY scheme improves the clustering effectiveness of QSS measures. To be able to apply the NN test, we designed a software test framework, Ferret, by complementing the features provided by dtSearch, a search engine. The test framework automates the generation of NN coefficients by processing standard test document collection data. We provide an insight into the construction and working of the Ferret test framework.

As a response to trade marks’ enhanced importance within trade, the EU’s interest in the area has increased by proponing a harmonization of the member states’ trade mark pro-tection so far as needed to preserve the EU’s objective of an internal market. The area is therefore regulated by an EU Directive, however allowing some national discretion. The purpose of this study was to investigate if a specific part of the trade mark protec-tion, the assessment-based confusion doctrine, corresponds on a Swedish and EU level. The aim was to locate statutory discrepancies in order to stimulate further review of the practical application of the doctrine from the analytical perspective of legal certainty. A scientifically accepted and traditional legal research method was applied when ex-amining and interpreting the sources of law. In addition, a comparative study was con-ducted between the two investigated legal systems to achieve the overall purpose. When comparing the results from the investigated sources, the legislations present a sta-tutory diversity, opening up for practical discrepancies. So was also the case with the application at the early stage of national implementation of the EU Directive. The tradi-tional national confusion doctrine, prescribing a more legal-technical assessment, did not correspond to the more flexible and contemporary EU view. Consequently, some national courts had to endure criticism for not adjusting to the EU development. Later case law however presents a very positive transition to the EU view of the confu-sion doctrine, suggesting a partial abandonment of the national legal sources of law for the benefit of EU law. Conclusion was however that despite this practical transition to EU law, statutory changes are necessary in order to safeguard the legal certainty in the way of achieving predictability.

The main contribution of this thesis, is to define similarity measures between two images with the main focus on content-based image retrieval (CBIR). Each image is considered as a set of visual elements that can be described with a set of visual descriptions (features). The similarity between images is then defined as the nearness between sets of elements based on a tolerance and a fuzzy tolerance relation. A tolerance relation is used to describe the approximate nature of the visual perception. A fuzzy tolerance relation is adopted to eliminate the need for a sharp threshold and hence model the gradual changes in perception of similarities. Three real valued similarity measures as well as a fuzzy valued similarity measure are proposed. All of the methods are then used in two CBIR experiments and the results are compared with classical measures of distance (namely, Kantorovich, Hausdorff and Mahalanobis). The results are compared with other published research papers. An important advantage of the proposed methods is shown to be their effectiveness in an unsupervised setting with no prior information. Eighteen different features (based on color, texture and edge) are used in all the experiments. A feature selection algorithm is also used to train the system in choosing a suboptimal set of visual features.

Automated map generalization is a necessary technique for the construction of multi-scale vector map databases that are crucial components in spatial data infrastructure of cities, provinces, and countries. Nevertheless, this is still a dream because many algorithms for map feature generalization are not parameter-free and therefore need human’s interference. One of the major reasons is that map generalization is a process of spatial similarity transformation in multi-scale map spaces; however, no theory can be found to support such kind of transformation. This thesis focuses on the theory of spatial similarity relations in multi-scale map spaces, aiming at proposing the approaches and models that can be used to automate some relevant algorithms in map generalization. After a systematic review of existing achievements including the definitions and features of similarity in various communities, a classification system of spatial similarity relations, and the calculation models of similarity relations in the communities of psychology, computer science, music, and geography, as well as a number of raster-based approaches for calculating similarity degrees between images, the thesis achieves the following innovative contributions. First, the fundamental issues of spatial similarity relations are explored, i.e. (1) a classification system is proposed that classifies the objects processed by map generalization algorithms into ten categories; (2) the Set Theory-based definitions of similarity, spatial similarity, and spatial similarity relation in multi-scale map spaces are given; (3) mathematical language-based descriptions of the features of spatial similarity relations in multi-scale map spaces are addressed; (4) the factors that affect human’s judgments of spatial similarity relations are proposed, and their weights are also obtained by psychological experiments; and (5) a classification system for spatial similarity relations in multi-scale map spaces is proposed. Second, the models that can calculate spatial similarity degrees for the ten types of objects in multi-scale map spaces are proposed, and their validity is tested by psychological experiments. If a map (or an individual object, or an object group) and its generalized counterpart are given, the models can be used to calculate the spatial similarity degrees between them. Third, the proposed models are used to solve problems in map generalization: (1) ten formulae are constructed that can calculate spatial similarity degrees by map scale changes in map generalization; (2) an approach based on spatial similarity degree is proposed that can determine when to terminate a map generalization system or an algorithm when it is executed to generalize objects on maps, which may fully automate some relevant algorithms and therefore improve the efficiency of map generalization; and (3) an approach is proposed to calculate the distance tolerance of the Douglas-Peucker Algorithm so that the Douglas-Peucker Algorithm may become fully automatic. Nevertheless, the theory and the approaches proposed in this study possess two limitations and needs further exploration. • More experiments should be done to improve the accuracy and adaptability of the proposed models and formulae. The new experiments should select more typical maps and map objects as samples, and find more subjects with different cultural backgrounds. • Whether it is feasible to integrate the ten models/formulae for calculating spatial similarity degrees into an identical model/formula needs further investigation. In addition, it is important to find out the other algorithms, like the Douglas-Peucker Algorithm, that are not parameter-free and closely related to spatial similarity relation, and explore the approaches to calculating the parameters used in these algorithms with the help of the models and formulae proposed in this thesis.

Every single day, lots of users actively participate in social media sites (e.g., Facebook, YouTube, Last.fm, Flicker, etc.) upload photos, videos, share bookmarks, write blogs and annotate/comment on content provided by others. With the recent proliferation of social media sites, users are overwhelmed by the huge amount of available content. Therefore, organizing and retrieving appropriate multimedia content is becoming an increasingly important and challenging task. This challenging task led a number of research communities to concentrate on social tagging systems (also known as folksonomy) that allow users to freely annotate their media items (e.g., music, images, or video) with any sort of arbitrary words, referred to as tags. Tags assist users to organize their own content, as well as to find relevant content shared by other users. In this thesis, we first analyze how useful a folksonomy is for improving personalized services such as tag recommendation, tag-based search and item annotation. We then propose two new algorithms for social media retrieval and tag recommendation respectively. The first algorithm computes the latent preferences of tags for users from other similar tags, as well as latent annotations of tags for items from other similar items. We then seamlessly map the tags onto items, depending on an individual user’s query, to find the most desirable content relevant to the user’s needs. The second algorithm improves tag-recommendation and item annotation by adapting the Katz measure, a path-ensemble based proximity measure, for the use in social tagging systems. In this algorithm we model folksonomy as a weighted, undirected tripartite graph. We then apply the Katz measure to this graph, and exploit it to provide personalized tag recommendation for individual users. We evaluate our algorithms on two real-world folksonomies collected from Last.fm and CiteULike. The experimental results demonstrate that the proposed algorithms improve the search and the recommendation performance, and obtain significant gains in cold start situations where relatively little information is known about a user or an item

The main contribution of this thesis, is to define similarity measures between two images with the main focus on content-based image retrieval (CBIR). Each image is considered as a set of visual elements that can be described with a set of visual descriptions (features). The similarity between images is then defined as the nearness between sets of elements based on a tolerance and a fuzzy tolerance relation. A tolerance relation is used to describe the approximate nature of the visual perception. A fuzzy tolerance relation is adopted to eliminate the need for a sharp threshold and hence model the gradual changes in perception of similarities. Three real valued similarity measures as well as a fuzzy valued similarity measure are proposed. All of the methods are then used in two CBIR experiments and the results are compared with classical measures of distance (namely, Kantorovich, Hausdorff and Mahalanobis). The results are compared with other published research papers. An important advantage of the proposed methods is shown to be their effectiveness in an unsupervised setting with no prior information. Eighteen different features (based on color, texture and edge) are used in all the experiments. A feature selection algorithm is also used to train the system in choosing a suboptimal set of visual features.

Submitted by Livia Mello (liviacmello@yahoo.com.br) on 2016-10-11T14:05:01Z No. of bitstreams: 1 DissGIF.pdf: 2443815 bytes, checksum: 100e28390f0b9b6780e2b2ce9aedfe39 (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-10-21T12:14:37Z (GMT) No. of bitstreams: 1 DissGIF.pdf: 2443815 bytes, checksum: 100e28390f0b9b6780e2b2ce9aedfe39 (MD5) / Approved for entry into archive by Marina Freitas (marinapf@ufscar.br) on 2016-10-21T12:14:43Z (GMT) No. of bitstreams: 1 DissGIF.pdf: 2443815 bytes, checksum: 100e28390f0b9b6780e2b2ce9aedfe39 (MD5) / Made available in DSpace on 2016-10-21T12:14:51Z (GMT). No. of bitstreams: 1 DissGIF.pdf: 2443815 bytes, checksum: 100e28390f0b9b6780e2b2ce9aedfe39 (MD5) Previous issue date: 2016-09-17 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / This paper presents an educational sequence for the Geometry classes, which explores the concept of similarity, focusing on the variation of the similarity ratio between linear measurement, areas measurements and volume of some polyhedrons. The problematic situation in this case involves a comparison between the prices of different popcorn packaging sizes in Bauru ́s movie theaters and its area. / Este trabalho apresenta uma sequência didática para aulas de Geometria abordando o conceito de semelhança com foco na variação da razão de semelhança entre medidas lineares e medidas de áreas e de volumes de alguns poliedros, tendo como situação problema a comparação de preços de diversos tamanhos de embalagens de pipocas vendidas em salas de cinema da região de Bauru.

Every single day, lots of users actively participate in social media sites (e.g., Facebook, YouTube, Last.fm, Flicker, etc.) upload photos, videos, share bookmarks, write blogs and annotate/comment on content provided by others. With the recent proliferation of social media sites, users are overwhelmed by the huge amount of available content. Therefore, organizing and retrieving appropriate multimedia content is becoming an increasingly important and challenging task. This challenging task led a number of research communities to concentrate on social tagging systems (also known as folksonomy) that allow users to freely annotate their media items (e.g., music, images, or video) with any sort of arbitrary words, referred to as tags. Tags assist users to organize their own content, as well as to find relevant content shared by other users. In this thesis, we first analyze how useful a folksonomy is for improving personalized services such as tag recommendation, tag-based search and item annotation. We then propose two new algorithms for social media retrieval and tag recommendation respectively. The first algorithm computes the latent preferences of tags for users from other similar tags, as well as latent annotations of tags for items from other similar items. We then seamlessly map the tags onto items, depending on an individual user’s query, to find the most desirable content relevant to the user’s needs. The second algorithm improves tag-recommendation and item annotation by adapting the Katz measure, a path-ensemble based proximity measure, for the use in social tagging systems. In this algorithm we model folksonomy as a weighted, undirected tripartite graph. We then apply the Katz measure to this graph, and exploit it to provide personalized tag recommendation for individual users. We evaluate our algorithms on two real-world folksonomies collected from Last.fm and CiteULike. The experimental results demonstrate that the proposed algorithms improve the search and the recommendation performance, and obtain significant gains in cold start situations where relatively little information is known about a user or an item

Labeling of trademark images with Vienna codes from the Vienna classification is a manual process carried out by domain experts, which enables searching trademark image databases using specific keywords that describe the semantic meaning of the figurative elements. In this research, we are investigating how application of supervised learning algorithms can improve and automate the manual process of labeling of new un-labeled trademark images. The successful implementation of deep learning algorithms in the task of computer vision for image classification has motivated us to investigate which of the supervised learning algorithms performs better trademark image classification. More specifically, to solve the problem of identification of figurative elements in new un-labeled images, we have used multi-class image classification approach based on deep learning and machine learning. To address this problem, we have generated a unique benchmarking dataset composed of 14,500 unique logos extracted from the European Union Intellectual Property Office Open Data Portal. The results after executing a set of controlled experiments on the given dataset indicate that deep learning models have overall better performance than machine learning models. In particular, CNN models reach better accuracy and precision, and significantly higher recall and F1 score for shorter training times, compared to recurrent neural networks such as LSTMs and GRUs. From the machine learning models, results indicate that Support Vector Machines have higher accuracy and overall better performance time compared to Decision Trees, Random Forests and Naïve Bayes models. This study shows that deep learning models can solve the problem of the labeling of trademark images with Vienna codes, and that can be applied by Intellectual Property Offices in real-world application for automation of the classification task which is carried out manually by the domain experts.

The ability to automate the process of comparing and matching resumes with job adverts is a growing research field. This can be done through the use of the machine learning area Natural Language Processing (NLP), which enables a model to learn human language. This thesis explores and evaluates the application of the state-of-the-art NLP model, SBERT, on the task of comparing and calculating a measure of similarity between extracted text from resumes and adverts. This thesis also investigates what type of data that generates the best performing model on said task. The results show that SBERT quickly can be trained on unlabeled data from the HR domain with the usage of a Triplet network, and achieves high performance and good results when tested on various tasks. The models are shown to be bilingual, can tackle unseen vocabulary and understand the concept and descriptive context of entire sentences instead of solely single words. Thus, the conclusion is that the models have a neat understanding of semantic similarity and relatedness. However, in some cases the models are also shown to become binary in their calculations of similarity between inputs. Moreover, it is hard to tune a model that is exhaustively comprehensive of such diverse domain such as HR. A model fine-tuned on clean and generic data extracted from adverts shows the overall best performance in terms of loss and consistency.