Master of Science / Department of Computing and Information Sciences / Doina Caragea / Over the last decade, the Internet has become an involving medium and user-generated content is continuously growing. Recommender systems that exploit user feedback are widely used in e-commerce and quite necessary for business enhancement. To make use of such user feedback, we propose a new content/collaborative hybrid approach, which is built on top of the recently released hetrec2011-movielens-2k dataset and is an extension of a previously proposed approach, called Weighted Tag Recommender (WTR). The WTR approach makes use of tag information available in hetrec2011-movielens-2k, but it does not use explicit ratings. As opposed to WTR, our modified approach can make use of ratings to capture collaborative filtering and either user-tags, available in the hetrec2011-movielens-2k, or movie keywords retrieved from IMDB, to capture movie content information. We call the two versions of our approach Weighted Tag Rating Recommender (WTRR) and Weighted Keyword Rating Recommender (WKRR), respectively. Movie keywords (which are not user specific) allow us to use all ratings available in hetrec2011-movielens-2k, as WKKR associates the content information from movies with the users, based on their ratings. On the other hand, tags provide more specific information for a user, but limit the usage of the data to the user-movie pairs that have tags (significantly smaller number compared with all pairs that have ratings). Both our keyword and tag representations of users can help alleviate the noise and semantic ambiguity problems inherent in information contributed by users of social networks. Experiments using the WTRR approach on a subset of the dataset (which contains both ratings and tags) show that it slightly outperforms the WKRR approach. However, WKRR can be applied to the whole hetrec2011-movielens-2k dataset and results show that the information from keywords can help build a movie recommender system competitive with other neighborhood based approaches and even with more sophisticated state-of-the-art approaches.

This thesis explores educator beliefs, attitudes and practices in teaching First Nations, Inuit, Métis and Ethnic students in Canada, including Muslim and Arabic students. Research supports evidence of on-going systemic racism toward the First Nations, Métis and Inuit, and New Canadians in society, and therefore, within our schools and our classrooms. Evidence also exists regarding how racism, invisibly instilled in educators as societal norms, permeates classrooms and helps continue to perpetuate marginalization and Othering of students. The research question was: How can teachers begin to deconstruct Euro-Centric, White Dominant beliefs, practices, attitudes and expectations in order to positively impact all students and their academic well-being? I examine the questions: What are some teachers doing which positively impacts learning for all students? What can be done to help teachers realize that some of their teaching practices and ways of dealing with students are disrespectful to the individual children and contribute to systemic racism?

This thesis is about keyword research for optimization of web pages. The thesis is mainly focused on on-page factors which are connected with keywords and key phrases. The goal of the thesis is design of methodics for keyword research and its evaluation on the basis of practical usage. The design of methodics is based on theoretical and practical knowledge and experience and it is the main contribution of the thesis. Usage of the keyword research is very wide, it forms the basis of creation of a web and content of the web. Everybody should learn the methodics, who wants to promote his web in competitive area, who wants to get to know his customers better and to generate higher profit, because the methodics is beneficial for all these and other reasons. The first part is theoretical basis. It shows the position of keyword research in SEO and in web mining. Further the important terms connected with keywords are defined. Keyword properties and methods to gaining keywords for website optimization are described either. Further various targets and metrics for measuring these targets are introduced. Next the most important tools for keyword research and characteristics of the two most used search engines for the Czech Republic (Google and Seznam) are described. The second part includes the design of the methodics for keyword research which consists of data collection, website optimization and monitoring of changes. Individual processes of methodis are described in practical experiments for particular web pages. The effectiveness of the methodics is evaluated at the end.

This thesis is aimed on design keyword detector. The work contains a description of the methods that are used for these purposes and design of algorithm for keyword detection. The proposed detector is based on the method of DTW (Dynamic Time Warping). Analysis of the problem was performed on the module programmed in ANSI C, which was created within the thesis. The results of the detector were evaluated using the metrics WER (word error rate) and AUC (area under curve).

No description available.

While news articles for print use one News Triangle, where important information is at the top of the article, online news articles are supposed to use a series of Stacked News Triangles, due to online readers text- skimming habits[1]. To identify Stacked News Triangles presence, we analyse how 68 Danish journalists annotate 31 articles. We use keyword frequency as the measure of popularity. To explore if Named Entities influence News Triangle presence, we analyse Named Entities found in the articles and keywords.We find the presence of an overall News Triangle in 30 of 31 articles, while, for the presence of Stacked News Triangles, 14 of the 31 articles have Stacked News Triangles. For Named Entities in News Triangles we cannot see what their influences is. Nonetheless, we find difference in Named Entity Types in each category (Culture, Domestic, Economy, Sports).

Nyckelord används på ett flertal sätt för att nå högre rankning i sökmotorresultaten på webben och många författare inom ämnet beskriver hur viktigt det är att använda rätt nyckelord när man sökop-timerar sin webbplats.I litteraturen om sökmotoroptimering beskrivs och kategoriseras nyckelord på ett flertal sätt. Teori-delen i denna uppsats är tänkt att vara en samlingsplats för information om nyckelord. Detta för att skapa en samlad helhetsbild av nyckelordsprocessen och beskriva de olika kategorierna av nyckelord så att detta finns dokumenterat på ett och samma ställe.I den andra delen av uppsatsen beskrivs en empirisk studie som gjorts för att få reda på om andra manuella metoder än den traditionella brainstormingen kan användas för att ta fram passande nyck-elord till en webbsida. De testade metoderna är slumpordsassociation, negativ idégenerering och 6-3-5. Även brainstorming testades för att jämföras med de andra tre metoderna. Våra resultat visar att negativ idégenerering verkar mindre lämplig för nyckelordsframtagning och att slumpordsassocia-tion och 6-3-5 är likvärdiga brainstorming och kan användas som framtagningsmetoder. / Keywords are used in different ways to achieve higher rankings in search engine results on the web and many authors in the field describe how important it is to use the right keywords when search optimizing a website.In the literature on search engine optimization, keywords are described and categorised in a number of ways. The literature study in this paper serves as a collection of information about keywords for web pages to gather the different categories and describe the keyword process so that this is docu-mented in one place.The second part of this paper describes an empirical study made to find out if other manual methods than the traditional brainstorming can be used to provide suitable keywords to a web page. The stu-died methods are random word association, negative idea generation and 6-3-5. Brainstorming was also tested so that it could be compared with the other three methods. Our results show that nega-tive idea generation seems less suitable for keyword generation and random word association and 6-3-5 are equivalent to brainstorming and can be used as generating methods.

Pas de résumé

Object category recognition is one of the most difficult problems in computer vision. It involves recognizing objects despite intra-class variations, viewpoint changes and background clutter. The goal of this thesis is to investigate robust invariant local image description and the selection of discriminative features. We show that class-discriminative scale-invariant features achieve excellent results for image-level categorization and object localization. We present solutions for two key problems: (i) we improve the quality of the image description based on a novel scale-invariant keypoint detection method and (ii) we integrate feature filtering techniques into our object models. Our novel scale-invariant detector is based on the idea of a ``maximally stable description'', i.e., the descriptor should be stable even in the presence of minor variations of the detector. The technique performs scale selection based on a region descriptor, here SIFT, and chooses regions for which this descriptor is maximally stable, i.e., the difference between descriptors extracted for consecutive scales reaches a minimum. This scale selection technique is applied to multi-scale Harris and Laplacian points. Experimental results evaluate the performance of our detector and show that it outperforms existing ones in the context of image matching, category and texture classification, as well as object localization. To construct object models based on discriminative features, we first cluster the scale-invariant descriptors and obtain a set of ``visual words''. We then estimate the discriminative information of these clusters based on different feature selection techniques---several of which are traditionally used in text retrieval. We discuss their properties---feature frequency, discriminative power, and redundancy---and analyze their performance in the context of image classification and object localization. We show that each task has different requirements, and indicate which selection techniques are the most appropriate. Experimental results for recognition on challenging large datasets demonstrate the performance of the approach.

QA bearing technologies Ltd. (hereafter QA) is a custom manufacturer of bearings in a variety of types and sizes. This research concerns one of the processes in QA’s bearing manufacturing in which the individual bearings, once assembled, are packaged to prevent corrosion and attack from any environmental influence. QA currently employs a manual process, which is time-consuming and significantly limits QA’s productivity. This greatly raises a need to improve this process through the design of an automated process for bearing packaging. Various kinds of packaging machine are available in the market for this application. However, most of them are designed for general packaging purpose and simply combining them may result in the functional overlap as applied to the packaging process. This research aims to design an automatic bearing packaging system (ABPS) based on an advanced design theory, i.e., Axiomatic Design Theory (ADT). In the first part of this research, a set of general functional requirements (FRs) and constraints (CONs) were defined specifically based on the needs from the customers. All FRs were maintained independent to each other such that the following design wouldn’t result in functional overlap. Then, design parameters (DPs) in general level were created to satisfy the defined FRs, each being a functionally independent subsystem consists of both custom-design components and products existing in the market. The second part of the research was to decompose the FRs and DPs based on independent axiom of ADT into further levels until the conceptual design was accomplished. In this process, acceptable alternatives in different hierarchies were proposed for comparison in the decision making process. Also, the relationship between FRs and DPs in the same level was always maintained one-to-one, which allows readily for modification in future, if needed. In the third part of this research, the information axiom of ADT was utilized to find the best solution from alternatives. For this, each alternative was evaluated in terms of the information content that reflects the possibility of satisfying the objective FR. The one with least information content was determined as the best alternative. As a result of this research, two solutions with different degree of automation were recommended for the bearing packaging system. The first solution is featured with higher degree of automation and capable of packaging bearings with limited human interface and as result, the total cost for its implementation is relatively high. The second solution integrates semi-automated machines and simplified sealing strategy and as such, the cost for its implementation is significantly reduced. After the conceptual design is complete, ADT is further applied in the detailed design level for oiling system with custom-designed components and mechanism. For the design of certain DPs, the design process applying ADT is used to compare with conventional methods, proved to be more efficient and simplified. Eventually, the final solution proposed for the detailed design of oiling system is consistent with independent axiom of ADT.