<div><div><div><p>As technological capacity to store and exchange information progress, the amount of available data grows explosively, which can lead to information overload. The dif- ficulty of making decisions effectively increases when one has too much information about that issue. Recommendation systems are a subclass of information filtering systems that aim to predict a user’s opinion or preference of topic or item, thereby providing personalized recommendations to users by exploiting historic data. They are widely used in e-commerce such as Amazon.com, online movie streaming com- panies such as Netflix, and social media networks such as Facebook. Memory-based collaborative filtering (CF) is one of the recommendation system methods used to predict a user’s rating or preference by exploring historic ratings, but without in- corporating any content information about users or items. Many studies have been conducted on memory-based CFs to improve prediction accuracy, but none of them have achieved better prediction accuracy than state-of-the-art model-based CFs. Fur- thermore, A product or service is not judged only by its own characteristics but also by the characteristics of other products or services offered concurrently. It can also be judged by anchoring based on users’ memories. Rating or satisfaction is viewed as a function of the discrepancy or contrast between expected and obtained outcomes documented as contrast effects. Thus, a rating given to an item by a user is a compar- ative opinion based on the user’s past experiences. Therefore, the score of ratings can be affected by the sequence and time of ratings. However, in traditional CFs, pairwise similarities measured between items do not consider time factors such as the sequence of rating, which could introduce biases caused by contrast effects. In this research, we proposed a new approach that combines both structural and rating-based similarity measurement used in memory-based CFs. We found that memory-based CF using combined similarity measurement can achieve better prediction accuracy than model-based CFs in terms of lower MAE and reduce memory and time by using less neighbors than traditional memory-based CFs on MovieLens and Netflix datasets. We also proposed techniques to reduce the biases caused by those user comparing, anchoring and adjustment behaviors by introducing the time-aware similarity measurements used in memory-based CFs. At last, we introduced novel techniques to identify, quantify, and visualize user preference dynamics and how it could be used in generating dynamic recommendation lists that fits each user’s current preferences.</p></div></div></div>

In a world of globalization, where offers continues to grow, the ability to direct people to their specific need is essential. After being key differentiating factors for Netflix and Amazon, Recommender Systems in general are no where near a downfall. Still, one downside of the basic recommender systems is that they are mainly based on indirect feedback (our behaviour, mainly form the past) as opposed to explicit demand at a specific time. Recent development in machine learning brings us closer to the possibility for a user to express it’s specific needs in natural language and get a machine generated reply. This is what Conversational Recommendation is about. Conversational recommendation encapsulates several machine learning sub-tasks. In this work, we focus our study on methods for the task of item (in our case, movie) recommendation from conversation. To explore this setting, we use, adapt and extend state of the art transformer based neural language modeling techniques to the task of recommendation from dialogue. We study the performance of different methods using the ReDial dataset [24], a conversational- recommendation dataset for movies. We also make use of a knowledge base of movies and measure their ability to improve performance for cold-start users, items, and/or both. This master thesis is divided as follows. First, we review all the basics concepts and the previous work necessary to to this lecture. When then dive deep into the specifics our data management, the different models we tested, the set-up of our experiments and the results we got. Follows the original a paper we submitted at RecSys 2020 Conference. Note that their is a minor inconsistency since throughout the thesis, we use v to represent items but in the paper, we used i. Overall, we find that pre-trained transformer models outperform baselines even if the baselines have access to the user preferences manually extracted from their utterances. / Dans un monde de mondialisation, où les offres continuent de croître, la capacité de référer les gens vers leurs besoins spécifiques est essentiel. Après avoir été un facteur de différenciation clé pour Netflix et Amazon, les systèmes de recommandation en général ne sont pas près de disparaître. Néanmoins, l’un des leurs inconvénients est qu’ils sont principalement basés sur des informations indirects (notre comportement, principalement du passé) par opposition à une demande explicite à un moment donné. Le développement récent de l’apprentissage automatique nous rapproche de la possibilité d’exprimer nos besoins spécifiques en langage naturel et d’obtenir une réponse générée par la machine. C’est ce en quoi consiste la recommandation conversationnelle. La recommandation conversationnelle englobe plusieurs sous-tâches d’apprentissage automatique. Dans ce travail, nous concentrons notre étude sur les méthodes entourant la tâche de recommandation d’item (dans notre cas, un film) à partir d’un dialogue. Pour explorer cette avenue, nous adaptons et étendons les techniques de modélisation du langage basées sur les transformeurs à la tâche de recommandation à partir du dialogue. Nous étudions les performances de différentes méthodes à l’aide de l’ensemble de données ReDial [24], un ensemble de données de recommandation conversationnelle pour les films. Nous utilisons également une base de connaissances de films et mesurons sa capacité à améliorer les performances lorsque peu d’information sur les utilisateurs/éléments est disponible. Ce mémoire par article est divisé comme suit. Tout d’abord, nous passons en revue tous les concepts de base et les travaux antérieurs nécessaires à cette lecture. Ensuite, nous élaborons les spécificités de notre gestion des données, les différents modèles que nous avons testés, la mise en place de nos expériences et les résultats que nous avons obtenus. Suit l’article original que nous avons soumis à la conférence RecSys 2020. Notez qu’il y a une incohérence mineure puisque tout au long du mémoire, nous utilisons v pour représenter les éléments mais dans l’article, nous avons utilisé i. Dans l’ensemble, nous constatons que les modèles de transformeurs pré-entraînés surpassent les modèles de bases même si les modèles de base ont accès aux préférences utilisateur extraites manuellement des dialogues.

Le travail de cette thèse s’inscrit dans le cadre de la recherche d’information (RI) et plus précisément la recherche d’information personnalisée. En effet, avec la prolifération des données de différentes sources et malgré la diversité de méthodes et d’approches de la recherche d’information classique, cette dernière n’est plus considérée aujourd’hui comme un moyen efficace pour répondre aux exigences de l’utilisateur considéré comme l’acteur principal de tout système de recherche d’information (SRI). Dans ce travail de thèse, nous adressons deux principaux problèmes liés à la RI personnalisée : (1) la formalisation et la mise en œuvre d’un modèle utilisateur et (2) la formulation de la requête de recherche dans le but d’améliorer les résultats retournés à l'utilisateur en fonction de sa perception et de ses préférences. Pour atteindre ces objectifs, nous avons proposé une approche de recherche d’information guidée par les ontologies et basée sur l’utilisation sémantique des informations. En effet, notre contribution se décline en trois principaux points : (1) la modélisation et la construction de profil utilisateur suivant une approche ontologique modulaire. Ce modèle permet de capturer les informations relatives à l’utilisateur, de les modéliser suivant l’approche sémantique dans le but de les réutiliser pour des tâches de raisonnement et d’inférence ; (2) la reformulation sémantique de la requête de recherche en exploitant les concepts, les relations syntaxiques et non syntaxiques entre les concepts et les propriétés ; et finalement, (3) la recommandation des résultats qui consiste à proposer des résultats de recherche en se basant sur l’ensemble des communautés utilisateurs construites par l’approche de classification non supervisée « Fuzzy K-mode » améliorée. Ces communautés sont aussi modélisées sémantiquement dans l’ontologie modulaire de profil. Ensuite, afin de valider l’approche proposée, nous avons mis en œuvre un système pour la recherche des itinéraires dans le transport public. Enfin, cette thèse propose des perspectives de recherche sur la base des limites rencontrées. / This PhD thesis reports on a recent study in the field of information retrieval (IR), more specifically personalized IR. Traditional IR uses various methods and approaches. However, given the proliferation of data from different sources, traditional IR is no longer considered to be an effective means of meeting users’ requirements. (‘Users’ here refers to the main actor in an IR system.) In this thesis, we address two main problems related to personalized IR: (1) the development and implementation of a user model; and (2) the formulation of a search query to improve the results returned to users according to their perceptions and preferences. To achieve these goals, we propose a semantic information search approach, based on the use of semantic information and guided by ontologies. The contribution of our work is threefold. First, it models and constructs user profiles following a modular ontological approach; this model allows the capture of information related to the user, and models the data according to the semantic approach so that the data can be re-used for reasoning and inference tasks. Second, it provides evidence for reformulating a query by exploiting concepts, hierarchical and non-hierarchical relationships between concepts and properties. Third, based on our findings, we recommend search results that are informed by the user’s communities, built by the improved unsupervised classification approach called the ‘Fuzzy K-mode’. These communities are also semantically modeled with modular profile ontology. To validate our proposed approach, we implemented a system for searching the itineraries for public transport. Finally, this thesis proposes research perspectives based on the limitations we encountered.

The Internet increases the demand for the development of commercial applications and services that can provide better shopping experiences for customers globally. It is full of information and knowledge sources that might confuse customers. This requires customers to spend additional time and effort when they are trying to find relevant information about specific topics or objects. Recommendation systems are considered to be an important method that solves this issue. Incorporating recommendation systems in online social networks led to a specific kind of recommendation system called social recommendation systems which have become popular with the global explosion in social media and online networks and they apply many prediction algorithms such as data mining techniques to address the problem of information overload and to analyze a vast amount of data. We believe that offering a real-time social recommendation system that can understand the real context of a user’s conversation dynamically is essential to defining and recommending interesting objects at the ideal time. In this thesis, we propose an architecture for a real-time social recommendation system that aims to improve word usage and understanding in social media platforms, advance the performance and accuracy of recommendations, and propose a possible solution to the user cold-start problem. Moreover, we aim to find out if the user’s social context can be used as an input source to offer personalized and improved recommendations that will help users to find valuable items immediately, without interrupting their conversation flow. The suggested architecture works as a third-party social recommendation system that could be incorporated with other existing social networking sites (e.g. Facebook and Twitter). The novelty of our approach is the dynamic understanding of the user-generated content, achieved by detecting topics from the user’s extracted dialogue and then matching them with an appropriate task as a recommendation. Topic extraction is done through a modified Latent Dirichlet Allocation topic modeling method. We also develop a social chat app as a proof of concept to validate our proposed architecture. The results of our proposed architecture offer promising gains in enhancing the real-time social recommendations.

The rise of disruptive technologies and the rapid growth of innovative initiatives have led to a trend of decentralization, deregulation, and distribution of regulated/centralized services. As a result, there is an increasing number of requests for the connection of distributed generators to distribution networks and the need for power utilities to quickly assess the impacts of distributed generators (DGs) to keep up with these requests. Grid integration of DGs brings about protection issues. Current protection systems were not designed for bi-directional power flow, thus the protective devices in the network lose their ability to perform their main functions. To mitigate the impact of distributed generation (DG), some standards and policies constrain the number of DG that can be connected to the distribution network. The problem with these limits is that they are based only on overload and overvoltage, and do not adequately define the DG size/threshold before the occurrence of a protection issue (NRS 097-2-3). The other problem with distributed generation is the vast difference in the technology, location, size, connection sequence, and protection scheme requirements which results in future DG network planning inadequacies – The Network DG Planning Dilemma. To determine the amount of DG to connect to the network, a detailed analysis is required which often involves the use of a simulation tool such as DIgSILENT to model the entire network and perform load flow studies. Modelling networks on DIgSILENT is relatively easy for simple networks but becomes time-consuming for complex, large, and real networks. This brings about a limitation to this method, planning inadequacies, and longer connection approval periods. Thus, there is a need for a fast but accurate system-wide tool that can assess the amount of DG that can be connected to a network. This research aims to present a technique used for calculating protection-based DG penetration limits on MV networks and develop a model to determine medium voltage opportunity network maps. These maps indicate the maximum amount of DG that can be connected to a network without the need for major protection scheme changes in South Africa. The approach to determining protection-based penetration limits is based on supervised machine learning methods. The aim is to rely on protection features present in the distribution network data i.e. fault level, Inverse Definite Minimum Time (IDMT) curve, pick-up current settings, Time Multiplier Settings (TMS), calculated relay operating times and relay positions to see how the network responds at certain DG penetration levels (‘actual' relay operating times). The dataset represents carefully anonymized distribution networks with accepted protection philosophy applied. A supervised machine learning algorithm is applied after nontrivial data pre-processing through recommendation systems and shuffling. The planning dilemma is cast into three parts: the first part is an automated pattern classification (logistic regression for classification of protection miscoordination), the second part involves regression (predicting operating time after different levels of DG penetration), and the last part involves developing a recommendation system (where, when and how much photovoltaic (PV) DG will be connected). Gradient descent, which is an optimisation algorithm that iterates and finds optimal values of the parameters that correspond to the local or global minimum values of the cost function using calculus was used to measure the accuracy of each model's hypothesis function. The cost function (one half mean squared error) for the models that predict ‘actual' relay operating times before DG penetration, at 35%, 65%, and 75% DG penetration converged to values below 120, 20, 15, and 15 seconds2 , respectively, within the first 100 iterations. A high variance problem was observed (cross-validation error was high and training error was low) for the models that used all the network protection features as inputs. The cross-validation and training errors approached the desired performance of 0.3±0.1 for the models that had second-order polynomials added. A training accuracy of 91.30%, 73.91%, 82.61%, and a validation accuracy of 100%, 55.56%, 66.67% was achieved when classifying loss of coordination, loss of grading and desensitization, respectively. A high bias problem was observed (cross-validation error was high and training error was high) for the loss of grading classification (relay positions eliminated) model. When the models (horizontal network features) were applied to four MV distribution networks, loss of coordination was not predicted, the loss of grading model had one false positive and the de-sensitization model had one false negative. However, when the results were compared to the vertical analysis (comparing the operating times of upstream and downstream relays/reclosers), 28 points indicated a loss of coordination (2 at 35%, 1 at 65% and 25 at 75% DG penetration). Protection coordination reinforcements (against loss of grading and desensitization) were found to be a requirement for DG connections where the MV transformer circuit breaker TMS is between 0.5 and 1.1, and where the network fault level is between 650 and 800A. Distribution networks in affluent neighbourhoods similar to those around the Western CapeSomerset West area and Gauteng- Centurion area need to be reinforced to accommodate maximum DG penetration up to the limit of 75% of the After Diversity Maximum Demand (ADMD). For future work, the collection of more data points (results from detailed analytical studies on the impact of DG on MV feeders) to use as training data to solve the observed high variance problem is recommended. Also, modifying the model by adding upstream and downstream network features as inputs in the classification model to solve the high bias problem is recommended.

No description available.

As society progresses towards a more algorithmic era, the influence of artificial intelligence (AI) is driving a revolution in the digital landscape. At its core, AI applications aim to engage customers by providing carefully tailored and data-driven personalization and customization of products, services, and marketing mix elements. However, the adoption of AI, while promising enhanced personalization, poses challenges due to the increased collection, analysis, and control of consumer data by technology owners. Consequently, concerns over data privacy have emerged as a primary consideration for individuals. This paper delves deeper into the implications of the personalization- privacy paradox, aiming to provide a comprehensive analysis of the challenges and opportunities it presents. The purpose of this thesis is to understand users’ privacy concerns and willingness to disclose their personal information to AI technologies by addressing the limitations of previous research and utilizing qualitative methods to gain a more in-depth understanding of consumer views. To understand users’ privacy concerns and willingness to disclose personal information to AI technologies, a qualitative approach was followed. Combining a deductive and inductive approach to fulfill the purpose of the study, empirical data was collected through 20 semi- structured interviews. The participants were chosen using a purposive sampling technique. Users’ privacy concerns and willingness to disclose personal information to AI technologies differ significantly. It depends not only on the individual, but also on the type of AI technology, the company providing the AI technology, the possibility of obtaining additional benefits, and whether the company is transparent about its data collection and can provide proof of security.

Bakgrund: Det förekommer stora nutritionsproblem på äldreboenden i Sverige och andra delar av världen. Sjukdom och åldrande är riskfaktorer för näringsmässiga störningar. Undernäring hos äldre är associerad med ökad dödlighet och sjuklighet samt en ökad risk för trycksårsutveckling och infektioner. Socialstyrelsen har utfärdat en rekommendation angående nattfastans längd för att motverka undernäring och dess komplikationer, samt för att öka energi- och näringsintaget hos vårdtagarna. Syfte: Att belysa nattfastan på äldreboenden i en stad i södra Sverige samt undersöka kunskap och attityder om nattfasta hos omvårdnadspersonalen på äldreboenden. Metod: Deskriptiv enkätundersökning med kvantitativ ansats. Studien gjordes på tolv äldreboenden där 167 informanter deltog. Urvalet bestod av omvårdnadspersonal från utvalda äldreboenden i en stad i södra Sverige. Datan analyserades i IBM SPSS Statistics version 22 och redovisades i deskriptiv statistik. Resultat: Majoriteten av informanterna hade vetskap om att Socialstyrelsen utfärdat en rekommendation gällande nattfasta. Dock var det knappt hälften som visste hur många timmar nattfastan ej bör överstiga. De flesta informanter uppgav att de erbjöd både förfrukost och ett extra mål på kvällen, men få vårdtagare nyttjar dessa mål. Slutsats: Många vårdtagare riskerar en alltför lång nattfasta då det är få som nyttjar de extra mål som erbjuds. Det är betydelsefullt att undervisa omvårdnadspersonal om de regelverk och rekommendationer som finns samt bedriva undervisning om nutritionsfrågor. Sjuksköterskan är ansvarig att se till att den kompetens som behövs finns. / Background: There are major nutrition problems in nursing homes in Sweden and other parts of the world. Illness and aging are risk factors for nutritional disorders. Malnutrition in the elderly is associated with increased morbidity and mortality and an increased risk of pressure ulceration and infection. The Swedish National Board has issued a recommendation regarding the length of night fasting to counter malnutrition and its complications and to increase energy and nutrient intake in the care recipients. Objective: To highlight the night fasting in nursing homes in a city in southern Sweden, and examine knowledge and attitudes about the night fasting in nursing staff in nursing homes. Method: Descriptive survey with quantitative approach. The study was done on twelve nursing homes where 167 informants participated. The sample consisted of nursing staff from selected nursing homes in a town in southern Sweden. The data were analysed in IBM SPSS Statistics version 22 and presented in descriptive statistics. Results: The majority of respondents had knowledge that the Swedish National Board issued a recommendation concerning night fasting. However, it was only nearly half who knew how many hours the night fasting should not exceed. Most informants stated that they offered both pre-breakfast and an extra meal in the evening, but few residents utilize these meals. Conclusion: Many residents run the risk of a long night fasting when there are few residents who use the additional snacks that offers. It is important to educate nursing staff on the rules and recommendations that exist and provide education on nutritional issues. The nurse is responsible to ensure that the skills needed are available.

Although they clearly exist, affinities among individuals are not all easily identified. Yet, they offer unique opportunities to discover new social networks, strengthen ties among individuals, and provide recommendations. We propose the idea of Implicit Affinity Networks (IANs) to build, visualize, and track affinities among groups of individuals. IANs are simple, interactive graphical representations that users may navigate to uncover interesting patterns. This thesis describes a system supporting the construction of IANs and evaluates it in the context of family history and online communities.

Recommendation system benefits social media by attracting users with the posts they prefer. The recommended posts, however, may not align with what users really need to browse, especially in terms of emotion. Thus we conducted a case study in TikTok, in order to understand the emotional impact of social application’s post feed and to explore the interactive solution. The state-of-arts were reviewed, on the topics of psychology issues caused by social media, related therapy and product solutions. To empathise with users’ situation, a workshop was performed, consisting of a card game, presentation and participatory design. Then an emotion reminder, built on a Naive Bayesian text classifier and a facial expression SVM, was prototyped. With an accuracy of 0.51 (text) and 0.69 (facial expression) in sentiment classification, the emotion reminder was then tested by the users. It was discovered that users had higher emotion awareness, higher sense of control over the browsing and lower engagement in the interface with the prototype, compared with the original TikTok interface. And this was aligned with their needs described in the workshop. Users preferred the prototype’s content-based emotion detection than the detection based on their biological data in terms of privacy, and embraced the format of the reminder, instead of auto-filter, as an emotionally positive experience was not just browsing the posts with positive feelings, but receiving negative posts as well. / Rekommendationssystem gynnar sociala medier genom att locka användare med de inlägg de föredrar. De rekommenderade inläggen kan dock inte alltid överensstämma med det användarna verkligen behöver bläddra igenom, särskilt när det gäller känslor. Därför genomförde vi en fallstudie på TikTok för att förstå den emotionella påverkan av sociala applikationers inläggflöde och för att utforska interaktiva lösningar. Den senaste forskningen inom området granskades med fokus på psykologiska problem orsakade av sociala medier, relaterad terapi och produktlösningar. För att sätta oss in i användarnas situation genomfördes en workshop med ett kortspel, presentation och deltagande design. Därefter skapades en känslomässig påminnelse, baserad på en Naive Bayes-textklassificerare och en SVM för ansiktsuttryck. Med en noggrannhet på 0,51 (text) och 0,69 (ansiktsuttryck) i känslolägesklassificering testades sedan känslominnaren av användarna. Det visade sig att användarna hade ökad medvetenhet om sina känslor, ökad känsla av kontroll över bläddrandet och lägre engagemang i gränssnittet med prototypen jämfört med det ursprungliga TikTok-gränssnittet. Detta stämde överens med deras behov som beskrevs under workshopen. Användarna föredrog prototypens innehållsbaserade känslodetektion jämfört med detektering baserad på deras biologiska data av integritetsskäl och omfamnade formatet på påminnelsen istället för automatisk filtrering. En emotionellt positiv upplevelse handlade inte bara om att bläddra bland inlägg med positiva känslor, utan även att ta emot negativa inlägg.