In this study we explore the problem of purchase cart recommendationin the field of retail. How can we push the right customize purchase cart that would consider both habits and serendipity constraints? Recommender Systems application is widely restricted to Internet services providers: movie recommendation, e-commerce, search engine. We brought algorithmic and technological breakthroughs to outdated retail systems while keeping in mind its own specificities: purchase cart rather than single products, restricted interactions between customers and products. After collecting ingenious recommendations methods, we defined two major directions - the correctness and the serendipity - that would serve as discriminant aspects to compare multiple solutions we implemented. We expect our solutions to have beneficial impacts on customers, gaining time and open-mindedness, and gradually obliterate the separation between supermarkets and e-commerce platforms as far as customized experience is concerned.

The recurrent neural networks have the ability to grasp the temporal patterns withinthe data. This is a property that can be used in order to help a recommender system bettertaking into account the user past history. Still the dimensionality problem that raiseswithin the recommender system field also raises here as the number of items the systemhave to be aware of is susceptibility high. Recent research have studied the use of such neural networks at a user’s session level.This thesis rather examines the use of this technique at a whole user’s past history levelassociated with techniques such as embeddings and softmax sampling in order to accommodatewith the high dimensionality. The proposed method results in a sequence prediction model that can be used as is forthe recommender task or as a feature within a more complex system. / De Recurrent Neural Networks har möjlighet att förstå de tidsmässiga mönstren inom data. Det här är en egenskap som kan användas för att hjälpa ett rekommendatörsystem bättre med hänsyn till användarens historia. Problemet med dimensioner inom rekommendatörsystem uppstår dock även här, eftersom antalet saker som systemet måste vara medveten om är extremt många. Nyare forskning har studerat användningen av sådana neurala nätverk på en användaressessionsnivå. Denna avhandling undersöker snarare användningen av denna teknik som en hel användares tidigare historiknivå i samband med tekniker som inbäddning och softmax-provtagning för att tillgodose den höga dimensionen. Den föreslagna metoden resulterar i en sekvensprediktionsmodell som kan användas som för recommender-uppgiften eller som en funktion inom ett mer komplext system.

Nowadays the largest E-commerce or E-service websites offer millions of products for sale. A Recommender system is defined as software used by such websites for recommending commercial or noncommercial product items to users according to the users’ tastes. In this project, we develop a recommender system for a private multimedia web service company. In particular, we devise three recommendation engines using different data filtering methods – named weighted-average, K-nearest neighbors, and item-based – which are based on collaborative filtering techniques, which work by recording user preferences on items and by anticipating the future likes and dislikes of users by comparing the records, for prediction of user preference. To acquire proper input data for the three engines, we retrieve data from database using three data collection techniques: active filtering, passive filtering, and item-based filtering. For experimental purpose we compare prediction accuracy of those three recommendation engines with the results from each engine and additionally we evaluate the performance of weighted-average method using an empirical analysis approach – a methodology which was devised for verification of predictive accuracy.

A cada dia aumenta o número de vídeos disponíveis no mundo. Por exemplo, há uma vasta quantidade de sites de vídeos disponíveis na Web e serviços de Vídeo Sob Demanda além de dispositivos que fazem a gravação de vídeos automaticamente, conhecidos como Personal Video Recorders, 24 horas por dia. Isso pode ocasionar um problema ao usuário: a sobrecarga de conteúdo em formato de vídeo. Uma das maneiras de se tratar tal problema consiste no uso de sistemas de recomendação, os quais filtram o conteúdo com o objetivo de entregar o que for mais interessante ao usuário. A abordagem típica utilizada pelos sistemas atuais consiste em um sistema de recomendação híbrido, i.e., que utiliza tanto filtragem baseada em conteúdo quanto filtragem colaborativa, minimizando os problemas que tais abordagens possuem individualmente. Adicionalmente, com o objetivo de melhorar a recomendação ou de criar novas formas de recomendação, têm sido apresentadas novas abordagens, tais como sistemas de recomendação utilizando dados de redes sociais, computação afetiva, tags, entre outros. Este trabalho tem como objetivo apresentar uma abordagem inovadora, a qual utiliza pontos de interesse em vídeo de usuários (ou seja, os segmentos dos vídeos que eles mais gostam ou que mais se interessam) para melhorar a acurácia de predição de sistemas de recomendação de vídeo que utilizam filtragem colaborativa baseados na abordagem usuário-usuário. Na abordagem proposta, os usuários participam de forma mais ativa e mais interativa ao marcarem seus pontos de interesse. Para avaliação de tal abordagem proposta foi realizada uma avaliação experimental em termos de acurácia de predição de avaliação; pela qual constatou-se que houve melhora na predição de avaliação do sistema de recomendação. Tal melhora está diretamente relacionada com o nível de participação das pessoas na marcação de pontos de interesse. / Every day the number of videos available in the world increases. For example, there is a vast amount of video sites available on the Web, Video On Demand services, as well as devices that records videos automatically, known as Personal Video Recorders, 24 hours a day. It may create a problem for the user: the overload of content in video format. One of the ways to treat such problem is the use of recommender systems, which filter the content in order to deliver what is most interesting to the user. The typical approach is to present a hybrid recommender system, i.e., that uses both contentbased filtering and collaborative filtering, minimizing the problems that these approaches have individually. Additionally, in order to improve the recommendation or to create new approaches of recommendation, has been given new approaches such as systems using data from social networks, affective computing, tags, etc. This paper aims to present an innovative approach, which uses points of interest (POI) in video of users (i.e., video segments best liked or most interested by them) to augment the prediction accuracy of video recommender systems with collaborative filtering based in the useruser approach. In the proposed approach, users participate more actively and more interactively to mark their points of interest. To evaluate this proposed approach an experimental evaluation was performed in terms of accuracy of ratings predictions; in which it was verified that there was an improvement in ratings prediction accuracy of the recommendation system. This improvement is directly related to the level of participation of people in marking points of interest.

A cada dia aumenta o número de vídeos disponíveis no mundo. Por exemplo, há uma vasta quantidade de sites de vídeos disponíveis na Web e serviços de Vídeo Sob Demanda além de dispositivos que fazem a gravação de vídeos automaticamente, conhecidos como Personal Video Recorders, 24 horas por dia. Isso pode ocasionar um problema ao usuário: a sobrecarga de conteúdo em formato de vídeo. Uma das maneiras de se tratar tal problema consiste no uso de sistemas de recomendação, os quais filtram o conteúdo com o objetivo de entregar o que for mais interessante ao usuário. A abordagem típica utilizada pelos sistemas atuais consiste em um sistema de recomendação híbrido, i.e., que utiliza tanto filtragem baseada em conteúdo quanto filtragem colaborativa, minimizando os problemas que tais abordagens possuem individualmente. Adicionalmente, com o objetivo de melhorar a recomendação ou de criar novas formas de recomendação, têm sido apresentadas novas abordagens, tais como sistemas de recomendação utilizando dados de redes sociais, computação afetiva, tags, entre outros. Este trabalho tem como objetivo apresentar uma abordagem inovadora, a qual utiliza pontos de interesse em vídeo de usuários (ou seja, os segmentos dos vídeos que eles mais gostam ou que mais se interessam) para melhorar a acurácia de predição de sistemas de recomendação de vídeo que utilizam filtragem colaborativa baseados na abordagem usuário-usuário. Na abordagem proposta, os usuários participam de forma mais ativa e mais interativa ao marcarem seus pontos de interesse. Para avaliação de tal abordagem proposta foi realizada uma avaliação experimental em termos de acurácia de predição de avaliação; pela qual constatou-se que houve melhora na predição de avaliação do sistema de recomendação. Tal melhora está diretamente relacionada com o nível de participação das pessoas na marcação de pontos de interesse. / Every day the number of videos available in the world increases. For example, there is a vast amount of video sites available on the Web, Video On Demand services, as well as devices that records videos automatically, known as Personal Video Recorders, 24 hours a day. It may create a problem for the user: the overload of content in video format. One of the ways to treat such problem is the use of recommender systems, which filter the content in order to deliver what is most interesting to the user. The typical approach is to present a hybrid recommender system, i.e., that uses both contentbased filtering and collaborative filtering, minimizing the problems that these approaches have individually. Additionally, in order to improve the recommendation or to create new approaches of recommendation, has been given new approaches such as systems using data from social networks, affective computing, tags, etc. This paper aims to present an innovative approach, which uses points of interest (POI) in video of users (i.e., video segments best liked or most interested by them) to augment the prediction accuracy of video recommender systems with collaborative filtering based in the useruser approach. In the proposed approach, users participate more actively and more interactively to mark their points of interest. To evaluate this proposed approach an experimental evaluation was performed in terms of accuracy of ratings predictions; in which it was verified that there was an improvement in ratings prediction accuracy of the recommendation system. This improvement is directly related to the level of participation of people in marking points of interest.

A cada dia aumenta o número de vídeos disponíveis no mundo. Por exemplo, há uma vasta quantidade de sites de vídeos disponíveis na Web e serviços de Vídeo Sob Demanda além de dispositivos que fazem a gravação de vídeos automaticamente, conhecidos como Personal Video Recorders, 24 horas por dia. Isso pode ocasionar um problema ao usuário: a sobrecarga de conteúdo em formato de vídeo. Uma das maneiras de se tratar tal problema consiste no uso de sistemas de recomendação, os quais filtram o conteúdo com o objetivo de entregar o que for mais interessante ao usuário. A abordagem típica utilizada pelos sistemas atuais consiste em um sistema de recomendação híbrido, i.e., que utiliza tanto filtragem baseada em conteúdo quanto filtragem colaborativa, minimizando os problemas que tais abordagens possuem individualmente. Adicionalmente, com o objetivo de melhorar a recomendação ou de criar novas formas de recomendação, têm sido apresentadas novas abordagens, tais como sistemas de recomendação utilizando dados de redes sociais, computação afetiva, tags, entre outros. Este trabalho tem como objetivo apresentar uma abordagem inovadora, a qual utiliza pontos de interesse em vídeo de usuários (ou seja, os segmentos dos vídeos que eles mais gostam ou que mais se interessam) para melhorar a acurácia de predição de sistemas de recomendação de vídeo que utilizam filtragem colaborativa baseados na abordagem usuário-usuário. Na abordagem proposta, os usuários participam de forma mais ativa e mais interativa ao marcarem seus pontos de interesse. Para avaliação de tal abordagem proposta foi realizada uma avaliação experimental em termos de acurácia de predição de avaliação; pela qual constatou-se que houve melhora na predição de avaliação do sistema de recomendação. Tal melhora está diretamente relacionada com o nível de participação das pessoas na marcação de pontos de interesse. / Every day the number of videos available in the world increases. For example, there is a vast amount of video sites available on the Web, Video On Demand services, as well as devices that records videos automatically, known as Personal Video Recorders, 24 hours a day. It may create a problem for the user: the overload of content in video format. One of the ways to treat such problem is the use of recommender systems, which filter the content in order to deliver what is most interesting to the user. The typical approach is to present a hybrid recommender system, i.e., that uses both contentbased filtering and collaborative filtering, minimizing the problems that these approaches have individually. Additionally, in order to improve the recommendation or to create new approaches of recommendation, has been given new approaches such as systems using data from social networks, affective computing, tags, etc. This paper aims to present an innovative approach, which uses points of interest (POI) in video of users (i.e., video segments best liked or most interested by them) to augment the prediction accuracy of video recommender systems with collaborative filtering based in the useruser approach. In the proposed approach, users participate more actively and more interactively to mark their points of interest. To evaluate this proposed approach an experimental evaluation was performed in terms of accuracy of ratings predictions; in which it was verified that there was an improvement in ratings prediction accuracy of the recommendation system. This improvement is directly related to the level of participation of people in marking points of interest.

Recommender systems support online customers by suggesting products and services of likely interest to them. Research in recommender systems is now starting to recognise the importance of multiple selection criteria and the role of customer context in improving the recommendation output. This thesis investigates the inclusion of criteria and context information in the recommendation process. Firstly, a novel technique for multi-criteria recommendation is proposed. It assumes that some selection criteria for an item (product or a service) will dominate the overall rating, and that these dominant criteria will be different for different users. Following this assumption, users are clustered based on their criteria preferences, creating a “preference lattice”. The recommendation output for a user is then based on ratings by other users from the same or nearby clusters. Secondly, a context similarity metric for context aware recommendation is presented. This metric can help improve the prediction accuracy in two ways. On the one hand, the metric can guide the aggregation of the feedback from similar context to improve the prediction accuracy. This aggregation is important because the recommendation generation based on prior feedback by similar customers reduces the quantum of feedback used, resulting in a reduction in recommendation quality. On the other hand, the value returned by the context similarity metric can also be used to indicate the importance of the context information in the prediction process for a context aware recommendation.The validation of the two proposed techniques and their applications are conducted in the service domain because the relatively high degree of user involvement attracts users to provide detailed feedback from multiple perspectives, such as from criteria and context perspectives. In particular, hotel services and web services areas are selected due to their different levels of maturity in terms of users’ feedback. For each area, this thesis proposes a different recommendation approach by combining the proposed techniques with a traditional recommendation approach. The thesis concludes with experiments conducted on the datasets from the two aforementioned areas to evaluate the proposed techniques, and to demonstrate the process and the effectiveness of the techniques-based recommendation approaches.

To combat fake news, researchers mostly focused on detecting fake news and journalists built and maintained fact-checking sites (e.g., Snopes.com and Politifact.com). However, fake news dissemination has been greatly promoted by social media sites, and these fact-checking sites have not been fully utilized. To overcome these problems and complement existing methods against fake news, in this thesis, we propose a deep-learning based fact-checking URL recommender system to mitigate impact of fake news in social media sites such as Twitter and Facebook. In particular, our proposed framework consists of a multi-relational attentive module and a heterogeneous graph attention network to learn complex/semantic relationship between user-URL pairs, user-user pairs, and URL-URL pairs. Extensive experiments on a real-world dataset show that our proposed framework outperforms seven state-of-the-art recommendation models, achieving at least 3~5.3% improvement.

Les systèmes de recommandation sont des outils servant à suggérer aux utilisateurs des items pouvant les intéresser. De tels systèmes requièrent la définition d'un algorithme prenant en compte le domaine d'application. Cet algorithme est ensuite exécuté pour chaque utilisateur du système afin de lui générer des recommandations, et ce, sans prendre en compte ses particularités et ses besoins spécifiques.L'objet de cette thèse consiste à proposer une nouvelle approche de recommandation hybride combinant plusieurs algorithmes de recommandation afin d'obtenir une recommandation plus précise. De plus, l'approche proposée repose sur la structure de l'ontologie donnée en entrée du système, ce qui la rend réutilisable, facilement adaptable et applicable à tous les domaines (musique, publications scientifiques, films, etc.).Nous nous sommes également intéressées à la détection du type de recommandations auxquelles l'utilisateur répond le mieux afin d'adapter le processus de recommandation à chaque catégorie d'utilisateur et d'obtenir des recommandations plus ciblées. Notre approche de recommandation permet également d'expliquer les recommandations obtenues, ce qui permet d'augmenter la confiance de l'utilisateur vis-à-vis du système en lui prouvant que ses recommandations lui sont personnellement destinées et de lui donner la possibilité de corriger les explications, ce qui améliore la connaissance de l'utilisateur par le système et aide à écarter les futures recommandations non pertinentes.Le système de recommandation défini a été expérimenté hors-ligne à l'aide d'une validation croisée sur le dataset de MovieLens et en ligne avec de vrais utilisateurs. Les résultats obtenus sont très satisfaisants. / Recommender systems are tools used to present users with items that might interest them. Such systems use algorithms that rely on the domain application. These algorithms are then executed for each user in order to find the most relevant recommendations for him, without taking into account his specific needs.In this thesis, we define a hybrid recommender system which combines several recommendation algorithms in order to obtain more accurate recommendations. Moreover, the defined approach relies on the structure of the input ontology, which makes the framework reusable, adaptable and domain-independent (music, research papers, films, etc.).We also had an interest in detecting in which kind of recommendations a user responds better in order to adapt the recommendation process to each user category and obtain more targeted recommendations. Finally, our approach can explain each recommendation, which increases the user confidence in the system by proving him that the recommendations are adapted to him. We also allow the user to correct the explanations in order to help the system to get a better understanding of him and avoid non accurate recommendations in the future.Our recommender system has been experimented online with real users and offline by performing a cross-validation on the MovieLens dataset. The results of the experimentation are very satisfying so far.

Consumer-generated reviews play an important role in online purchase decisions for many consumers. However, the quality and helpfulness of online reviews varies significantly. In addition, the helpfulness of different consumer-generated reviews is not disclosed to consumers unless they carefully analyze the overwhelming number of available contents. Therefore, it is of vital importance to develop predictive models that can evaluate online product reviews efficiently and then display the most useful reviews to consumers, in order to assist them in making purchase decisions. This thesis examines the problem of building computational models for predicting whether a consumer-generated review is helpful based on consumers' online votes on other reviews (where a consumer's vote on a review is either HELPFUL or UNHELPFUL), with the aim of suggesting the most suitable products and vendors to consumers.In particular, we propose in this thesis three different helpfulness prediction approaches for consumer-generated reviews. Our entropy-based approach is relatively simple and suitable for applications requiring simple recommendation engine with fully-voted reviews. However, our entropy-based approach, as well as the existing approaches, lack a general framework and are all limited to utilizing fully-voted reviews. We therefore present a probabilistic helpfulness prediction framework to overcome these limitations. To demonstrate the versatility and flexibility of this framework, we propose an EM-based model and a logistic regression-based model. We show that the EM-based model can utilize reviews voted by a very small number of voters as the training set, and the logistic regression-based model is suitable for real-time helpfulness predicting of consumer-generated reviews. To our best knowledge, this is the first framework for modeling review helpfulness and measuring the goodness of models. Although this thesis primarily considers the problem of review helpfulness prediction, the presented probabilistic methodologies are, in general, applicable for developing recommender systems that make recommendation based on other forms of user-generated contents.