Cette thèse est consacrée à l’étude des systèmes de recommandation basés sur des réseaux de neurones artificiels appris pour faire de l'ordonnancement de produits avec des retours implicites (sous forme de clics). Dans ce sens, nous proposons un nouveau modèle neuronal qui apprend conjointement la représentation des utilisateurs et des produits dans un espace latent, ainsi que la relation de préférence des utilisateurs sur les produits. Nous montrons que le modèle proposé est apprenable au sens du principe de la minimisation du risque empirique et performant par rapport aux autres modèles de l'état de l'art sur plusieurs collections. En outre, nous contribuons à la création de deux nouvelles collections, produites grâce aux enregistrements des comportements de clients de Kelkoo (https://www.kelkoo.com/); le leader européen de la publicité programmatique et de Purch (http://www.purch.com/). Les deux jeux de données recueillent des retours implicites des utilisateurs sur des produits, ainsi qu’un grand nombre d'informations contextuelles concernant à la fois les clients et les produits. La collections de données de Purch contient en plus une information sur la popularité des produits ainsi que des commentaires textuelles associés. Nous proposons, une stratégie simple et efficace sur la manière de prendre en compte le biais de la popularité ainsi qu'un modèle probabiliste latent temporel pour extraire automatiquement les thèmes des textes des commentaires.Mots clés. Systèmes de recommandation, apprentissage d'ordonnancement, réseaux de neurones, recommandations avec des retours implicites, Modèles probabilistes latents temporels / This thesis is dedicated to the study of Recommendation Systems for implicit feedback (clicks) mostly using Learning-to-rank and neural network based approaches. In this line, we derive a novel Neural-Network model that jointly learns a new representation of users and items in an embedded space as well as the preference relation of users over the pairs of items and give theoretical analysis. In addition we contribute to the creation of two novel, publicly available, collections for recommendations that record the behavior of customers of European Leaders in eCommerce advertising, Kelkoofootnote{url{https://www.kelkoo.com/}} and Purchfootnote{label{purch}url{http://www.purch.com/}}. Both datasets gather implicit feedback, in form of clicks, of users, along with a rich set of contextual features regarding both customers and offers. Purch's dataset, is affected by popularity bias. Therefore, we propose a simple yet effective strategy on how to overcome the popularity bias introduced while designing an efficient and scalable recommendation algorithm by introducing diversity based on an appropriate representation of items. Further, this collection contains contextual information about offers in form of text. We make use of this textual information in novel time-aware topic models and show the use of topics as contextual information in Factorization Machines that improves performance. In this vein and in conjunction with a detailed description of the datasets, we show the performance of six state-of-the-art recommender models.Keywords. Recommendation Systems, Data Sets, Learning-to-Rank, Neural Network, Popularity Bias, Diverse Recommendations, Contextual information, Topic Model.

A anemia ferropriva no Brasil é o problema carencial de maior magnitude tendo como principal fator a insuficiência de alimentos fontes de ferro na dieta. O objetivo deste estudo foi analisar, por meio de um exercício teórico, a capacidade da alimentação atender à recomendação e necessidade de ferro levando em consideração a legislação para fortificação das farinhas de trigo e de milho com o mineral vigente no Brasil desde junho de 2004. Para esse exercício foram utilizados os dados secundários referidos em relação à aquisição alimentar domiciliar pela população brasileira na Pesquisa de Orçamentos Familiares 2002/2003, destacando as macrorregiões e, entre as classes de rendimento, exclusivamente os dados referentes às famílias com rendimento mensal inferior a dois salários mínimos. Os valores encontrados foram organizados para estimar a disponibilidade e biodisponibilidade do ferro na alimentação adquirida e realizar o cálculo da densidade de ferro e densidade de ferro biodisponível. A partir da aquisição alimentar domiciliar verificou-se um valor calórico diário insuficiente para atender à necessidade energética. Verificou-se que o baixo teor de ferro proveniente dos alimentos naturalmente fontes do mineral feijão e carnes foi acrescido em 45% com o ferro dos alimentos derivados das farinhas de trigo e de milho fortificadas. Entre as famílias com renda mensal inferior a dois salários mínimos o aumento foi de 40%. Mesmo com a fortificação, a quantidade de ferro veiculada pela alimentação nacional está longe de atingir a recomendação de 14mg Fe/dia. O valor encontrado atende a recomendação marcial para o homem, mas representa apenas 55% do recomendado para a mulher. A densidade do ferro biodisponível de 0,360mg/1000 Kcal também não atendeu à necessidade diária do homem (1,0 mg/dia) e muito menos da mulher (2,2 mg/dia). A elevada prevalência com que essa deficiência ocorre, justificada pelo baixo consumo do mineral, acarreta à Saúde Pública elevados custos diretos e indiretos. A fortificação das farinhas com ferro é uma medida destinada a grandes segmentos populacionais e visa principalmente à prevenção e controle da deficiência marcial. / The iron deficiency anemia in Brazil is a fundamental problem of greater magnitude, and its main factor is that in the Brazilian diet there is insufficient use of foods that provide sources of iron. The objective of this piece of work is, by means of a theoretical work, to analyze if the diet is able to meet the recommendation / need of iron taking into account the legislation of fortification of wheat flour and corn with the current mineral in Brazil since June, 2004. For this matter, secondary data referring the domestic food acquisition by the Brazilian population in the \"Pesquisa de Orçamentos Familiares\" household budget survey 2002 / 2003 were used, focusing on the biggest regions and the income classes, especially the ones with monthly income inferior to two minimal salaries. The values found were organized to estimate the availability and the bioavailability of iron in the diet and to calculate the density of iron and of bioavailable iron. Based on the domestic food acquisition, a caloric value, which was insufficient to meet the needs of energy consumption, was observed. It was also noticed that the low amount of iron coming from foods naturally found in beans and meat was increased by 45% with the iron derived from fortified wheat flour and corn. Among families whose income was inferior to two minimal salaries, the increase was 40%. Even with the fortification, the quantity of iron available in the usual diet of the country is far behind the recommended 14mg per day. The amount meets the recommended for male, but it only adds up to 55% of what is recommended for the females. The density of boiavailable iron, which was 0.360 mg / 1000kcal, was also unable to meet the daily need of a male (1.0mg per day) and most notably of a female (2.2mg per day). The high prevalence that this deficiency happens, justify by the low consumption of this mineral, brings very high direct and indirect expenses to the Brazilian Department of Public Health Public. The fortification of flour with iron is a measure destined to great population segments and is aimed specifically at the prevention and control of martial deficiency.

Marijuana is the most frequently used illegal substance in the United States and is most widely used among young people aged 12 to 21 years. Accurate screening and monitored issuance of medical cannabis recommendations have been shown to decrease abuse rates of the substance, create fewer deaths from opiates, reduce crime rates, reduce marijuana use in youths, decrease car crash deaths, and lessen prevalence of suicide in young men. The purpose of this project was to explore whether multiple screening methods for depression and anxiety in patients who seek medical cannabis referrals for anxiety and depression would improve screening and cannabis referral accuracy. A comprehensive review of the literature was conducted, and 2 screening tools were identified. The tools identified were the Zung self-rating anxiety scale and the Diagnostic and Statistical Manual of Mental Disorders 5 scale. The medical director at the project site reviewed the tools and approved them. These tools were then included in an education program for 12 staff members and providers with a pretest given to the participants prior to the staff education program. A posttest was then administered to the same group after the staff education program was completed and the new screening measures implemented. Results showed that referrals for cannabis were at 85% before the 2-step screening process was implemented; referrals for cannabis decreased to 60% with implementation of the dual screening method, suggesting increased accuracy in screening for depression and anxiety for cannabis referrals. This project might promote positive social change by increasing accuracy for cannabis referrals and reducing the risk of cannabis abuse.

The primary purpose of this study is to explore a case of remarkably powerful contemporary rhetoric, namely Robert Parker’s wine writing, which has had an unprecedented impact in the world of prestigious wine for more than two decades. Parker, an American autodidact who gave up his career in law to become a fulltime wine critic, is considered the most influential critic of all time. This background motivates the approach of the current enquiry, which targets the persuasiveness in Parker’s writing. The investigation strives to bring to the fore both explicit and implicit elements of his wine reviews that have the potential to contribute to rhetorical success. The material selected for analysis comprises a corpus of reviews extracted from Parker’s extensive bulk of wine writing. The texts are studied against the backdrop of socio-cultural and institutional frames. Considerable importance is assigned to the fact that the reviews occur within a strictly specialized field of discourse with a highly conventionalized configuration. This hermeneutic enquiry approaches the topic from three analytical perspectives, designed to highlight persuasiveness in representations, argumentation and appraisal. The presentation reports on schematic patterns in Parker’s discourse as well as close interpretation of individual texts. The analysis of representations shows that both visual and verbal representations contribute to the persuasiveness of the text. The argumentative exploration of Parker’s discourse, which is assisted by the analytical tools of pragma-dialectics, demonstrates that the reviews involve rational argumentation on several subordinate levels, given in support of assessments and recommendations. Finally, the perspective of appraisal draws on the analytical resources provided by the Appraisal model to shed light on the way in which the audience is positioned to respond with respect to emotional, associative and perceptual values. The results indicate that the persuasiveness of Parker’s discourse arises as a result of concordance among an intricate array of interrelated factors. The audience is recurrently demonstrated to play a crucial role as co-constructors of the message. The present study also has methodological outcomes, presenting a novel combination of analytical methods to perform contextually situated discourse analysis. In addition, the material is allowed to challenge the theoretical ideas and notions that are addressed.

Contrary to popular belief, the L.A. region is more of a desert than a tropical oasis. Little rainfall during the winter months and practically no rainfall during the summer months is characteristic of Southern California’s desert-like weather patterns. Due to these low precipitation levels, water is considered the most important commodity in the Los Angeles region. Prior to 1900, the inhabitants of this area were fully aware of the importance of water. Most settlements were established near water sources and had adopted various techniques and constructed small-scale dams to conserve and reuse rainwater. Yet these measures were not sufficient to sustain large populations during drought conditions. Most settlers were forced to seek other areas where more reliable sources of water were found. The construction of early engineering feats like the Los Angeles aqueduct quickly changed prior perceptions of the region’s potential. Such systems allowed for cheap and previously inaccessible water to flow to the abundant land, spurring an unprecedented population and agricultural boom. For decades, the construction of more aqueducts and canals provided a sufficient amount of water to meet the demand in the region’s growing agricultural and financial economy. As the abundance of land and favorable weather attracted more businesses and industries into the region, more and more homes were built to accommodate the workforce. By 1936, the Hoover Dam had been built and California had signed and agreed to the Colorado River Compact, which granted Southern California 4.4 million acre-feet annually of the Colorado River’s water. As a result, relatively cheap water was able to meet the demands, thus catapulting California’s agricultural industry and residential development.

Facebook自從在台灣推出以來，已有超過一千三百萬的使用者帳號，是最熱門的社群網站，其中蘊含了龐大的使用者資料。從使用者學歷、工作經歷和喜歡的粉絲頁中可以一定程度上地判斷出使用者的背景與喜好，若能利用分析過的資訊將使用者分群，以供交友或導向到可能喜歡的粉絲頁，就能開發潛在客戶進而掌握商機。 本研究旨在完成一個線上系統，透過Facebook上可供擷取個人的資料：學歷、工作經歷以及喜歡的粉絲頁等資訊，針對這些量化過的資訊，經Kmeans將使用者分群分類，藉以作為協同過濾式推薦。目前實驗結果將有效個人資料4417筆進行分群，以使用者喜歡的粉絲頁比例(本研究整合成48種)加上工作經歷與學歷，最終分成10群，以作為交叉推薦之憑據和延伸研究。研究過程分實驗組與對照組，實驗組是本研究推薦的10筆粉絲頁，也就是使用者與所屬群集質心比例相差較多的粉絲頁類型；對照組則是選取使用者與母體中有較多比例差距的10筆，以證明本研究的推薦模型有效。 最後由使用者針對兩組推薦結果進行滿意度評分之比較，總共收回使用者回饋68筆，實驗組與對照組的平均推薦滿意度分數分別為0.5743、0.4268，對兩者作信心水準為95%的t檢定，結果為有充分證據支持實驗組大於對照組，可證明本研究對於推薦準確性的幫助，達成本研究目的。 由此實驗可以確定在Facebook上以使用者屬性為基礎的粉絲頁與人脈推薦是有意義與價值的，也說明真實數據能應用在社群網站的研究。希冀本研究的結果能帶動其他社群網站研究朝使用真實數據去分析佐證，讓社群網站的研究結果能更貼近使用者的真實行為。 / Facebook is one of the most popular social websites in Taiwan, and it has over 13 million accounts with lots of user data. One can tell a user’s background and preference by his education, work experience, and preferred fan pages. If we direct the right user to the right fan pages by analyzing information and clustering users through recommendation or personal connections, we will be able to reach potential customers and to further business opportunities. The goal of this study is to complete an online system to assume collaborative fan page recommendation. Base on users’ education degree, work experience and preferred fan pages, users’ background. Then use the Kmeans algorithm to cluster quantified personal information to recommend fan pages and social relationships. Currently, the result of the experiment shows 10 clusters, which contain 4417 users, and we use it as a foundation of crossing recommendation. To prove the effect of this study, we divide study into two groups, an experimental group and control group. The former one represents recommended top 10 fan pages that include the fan page types with highest difference of percentage between user’s attributes and cluster centroid; the latter one represents top 10 fan pages that include the fan page types with highest difference of percentage between users’ attributes and proportion respectively. Finally, we use users score satisfaction for each group to compare. There are 68 pieces of feedback, and the average satisfaction scores of the experimental group and the control group are 0.5743 and 0.4268, respectively. On both a confidence level of 95% for t-test, the result shows there is more sufficient evidence to support the satisfaction of experimental group than the control group. We can prove accuracy for recommendation to achieve the goal in this study. This experiment determines not only the fan page recommendation based on user attributes on Facebook is meaningful and valuable, but also shows real data can be used in social networking studies. We hope the results of this study can lead other social networking studies to analyze with real users’ data in order to make future study on social networking better reflect real users’ behavior.

Given set of web pages labeled with ontological items, the level of similarity between two web pages is measured using the level of similarity between ontological items of pages labeled with. Using similarity measure between two pages, degree of similarity between two sequences of web page visits can be calculated as well. Using clustering algorithms, similar frequent sequences are grouped and representative sequences are selected from these groups. A new sequence is compared with all clusters and it is assigned to most similar one. Representatives of the most similar cluster can be used in several real world cases. They can be used for predicting and prefetching the next page user will visit or for helping the navigation of user in the website. They can also be used to improve the structure of website for easier navigation. In this study the effect of time spent on each web page during the session is analyzed.

<p>This master thesis has been written at the School of Mathematics and System Engineering (MSI) at Växjö University in the field of computer science. IT security has become one of the main topics of every enterprise since they all use information technology in their business. Investments have to be done in order to achieve a high protection status of the IT environments. Specially small and medium-sized enterprises need more knowledge and advisory how to handle their IT security. This thesis analyses management tools that have the goal to improve the IT security. The second and main part of the thesis is the design of a tool that should helps to solve the described problems.</p>

Recommendation systems are emerging as an important business application as the demand for personalized services in E-commerce increases. Collaborative filtering techniques are widely used for predicting a user's preference or generating a list of items to be recommended. In this thesis, we develop several new approaches for collaborative filtering based on model combination and kernel smoothing. Specifically, we start with an experimental study that compares a wide variety of CF methods under different conditions. Based on this study, we formulate a combination model similar to boosting but where the combination coefficients are functions rather than constant. In another contribution we formulate and analyze a local variation of matrix factorization. This formulation constructs multiple local matrix factorization models and then combines them into a global model. This formulation is based on the local low-rank assumption, a slightly different but more plausible assumption about the rating matrix. We apply this assumption to both rating prediction and ranking problems, with both empirical validations and theoretical analysis. We contribute with this thesis in four aspects. First, the local approaches we present significantly improve the accuracy of recommendations both in rating prediction and ranking problems. Second, with the more realistic local low-rank assumption, we fundamentally change the underlying assumption for matrix factorization-based recommendation systems. Third, we present highly efficient and scalable algorithms which take advantage of parallelism, suited for recent large scale datasets. Lastly, we provide an open source software implementing the local approaches in this thesis as well as many other recent recommendation algorithms, which can be used both in research and production.

Internet Protocol TV (IPTV) delivers television content to users over IP-based network. Different from the traditional TV services, IPTV platforms provide users with large amount of multimedia contents with interactive and personalized services, including the targeted advertisement, on-demand content, personal video recorder, and so on. IPTV is promising since it allows to satisfy users experience and presents advanced entertainment services. On the other hand, the Next Generation Network (NGN) approach in allowing services convergence (through for instance coupling IPTV with the IP Multimedia Subsystem (IMS) architecture or NGN Non-IMS architecture) enhances users' experience and allows for more services personalization. Although the rapid advancement in interactive TV technology (including IPTV and NGN technologies), services personalization is still in its infancy, lacking the real distinguish of each user in a unique manner, the consideration of the context of the user (who is this user, what is his preferences, his regional area, location, ..) and his environment (characteristics of the users' devices 'screen types, size, supported resolution, '' and networks available network types to be used by the user, available bandwidth, ..') as well as the context of the service itself (content type and description, available format 'HD/SD', available language, ..) in order to provide the adequate personalized content for each user. This advanced IPTV services allows services providers to promote new services and open new business opportunities and allows network operators to make better utilization of network resources through adapting the delivered content according to the available bandwidth and to better meet the QoE (Quality of Experience) of clients. This thesis focuses on enhanced personalization for IPTV services following a user-centric context-aware approach through providing solutions for: i) Users' identification during IPTV service access through a unique and fine-grained manner (different from the identification of the subscription which is the usual current case) based on employing a personal identifier for each user which is a part of the user context information. ii) Context-Aware IPTV service through proposing a context-aware system on top of the IPTV architecture for gathering in a dynamic and real-time manner the different context information related to the user, devices, network and service. The context information is gathered throughout the whole IPTV delivery chain considering the user domain, network provider domain, and service/content provider domain. The proposed context-aware system allows monitoring user's environment (devices and networks status), interpreting user's requirements and making the user's interaction with the TV system dynamic and transparent. iii) Personalized recommendation and selection of IPTV content based on the different context information gathered and the personalization decision taken by the context-aware system (different from the current recommendation approach mainly based on matching content to users' preferences) which in turn highly improves the users' Quality of Experience (QoE) and enriching the offers of IPTV services