Master of Science / Department of Computing and Information Sciences / Doina Caragea / With the rising trend in social networking, more people express their opinions on the web. As a consequence, there has been an increase in the number of blogs where people write reviews about the products they buy or services they experience. These reviews can be very helpful to other potential customers who want to know the pros and cons of a product, and also to manufacturers who want to get feedback from customers about their products. Sentiment analysis of online data (such as review blogs) is a rapidly growing field of research in Machine Learning, which can leverage online reviews and quickly extract the sentiment of a whole blog. The accuracy of a sentiment analyzer relies heavily on correctly identifying associations between a sentiment (opinion) word and the targeted mention (token or object) in blog sentences. In this work, we focus on the task of automatically identifying sentiment-mention associations, in other words, we identify the target mention that is associated with a sentiment word in a sentence. Support Vector Machines (SVM), a supervised machine learning algorithm, was used to learn classifiers for this task. Syntactic and semantic features extracted from sentences were used as input to the SVM algorithm. The dataset used in the work has reviews from car and camera domain. The work is divided into two phases. In the first phase, we learned domain specific classifiers for the car and camera domains, respectively. To further improve the predictions of the domain specific classifiers we investigated the use of transfer learning techniques in the second phase. More precisely, the goal was to use knowledge from a source domain to improve predictions for a target domain. We considered two transfer learning approaches: a feature level fusion approach and a classifier level fusion approach. Experimental results show that transfer learning can help to improve the predictions made using the domain specific classifier approach. While both the feature level and classifier level fusion approaches were shown to improve the prediction accuracy, the classifier level fusion approach gave better results.

A análise de sentimentos é um campo de estudo com recente popularização devido ao crescimento da Internet e do conteúdo que é gerado por seus usuários, principalmente nas redes sociais, nas quais as pessoas publicam suas opiniões em uma linguagem coloquial e em muitos casos utilizando de artifícios gráficos para tornar ainda mais sucintos seus diálogos. Esse cenário é observado no Twitter, uma ferramenta de comunicação que pode facilmente ser usada como fonte de informação para várias ferramentas automáticas de inferência de sentimentos. Esforços de pesquisas têm sido direcionados para tratar o problema de análise de sentimentos em redes sociais sob o ponto de vista de um problema de classificação, com pouco consenso sobre qual é o classificador com melhor poder preditivo, bem como qual é a configuração fornecida pela engenharia de atributos que melhor representa os textos. Outro problema é que em um cenário supervisionado, para a etapa de treinamento do modelo de classificação, é imprescindível se dispor de exemplos rotulados, uma tarefa árdua e que demanda esforço humano em grande parte das aplicações. Esta tese tem por objetivo investigar o uso de agregadores de classificadores (classifier ensembles), explorando a diversidade e a potencialidade de várias abordagens supervisionadas quando estas atuam em conjunto, além de um estudo detalhado da fase que antecede a escolha do classificador, a qual é conhecida como engenharia de atributos. Além destes aspectos, um estudo mostrando que o aprendizado não supervisionado pode fornecer restrições complementares úteis para melhorar a capacidade de generalização de classificadores de sentimento é realizado, fornecendo evidências de que ganhos já observados em outras áreas do conhecimento também podem ser obtidos no domínio em questão. A partir dos promissores resultados experimentais obtidos no cenário de aprendizado supervisionado, alavancados pelo uso de técnicas não supervisionadas, um algoritmo existente, denominado de C3E (Consensus between Classification and Clustering Ensembles) foi adaptado e estendido para o cenário semissupervisionado. Este algoritmo refina a classificação de sentimentos a partir de informações adicionais providas pelo agrupamento em um procedimento de autotreinamento (self-training). Tal abordagem apresenta resultados promissores e competitivos com abordagens que representam o estado da arte em outros domínios. / Sentiment analysis is a field of study that shows recent popularization due to the growth of Internet and the content that is generated by its users. More recently, social networks have emerged, where people post their opinions in colloquial and compact language. This is what happens in Twitter, a communication tool that can easily be used as a source of information for various automatic tools of sentiment inference. Research efforts have been directed to deal with the problem of sentiment analysis in social networks from the point of view of a classification problem, where there is no consensus about what is the best classifier, and what is the best configuration provided by the feature engineering process. Another problem is that in a supervised setting, for the training stage of the classification model, we need labeled examples, which are hard to get in the most of applications. The objective of this thesis is to investigate the use of classifier ensembles, exploring the diversity and the potential of various supervised approaches when these work together, as well as to provide a study about the phase that precedes the choice of the classifier, which is known as feature engineering. In addition to these aspects, a study showing that unsupervised learning techniques can provide useful and additional constraints to improve the ability of generalization of the classifiers is also carried out. Based on the promising results got in supervised learning settings, an existing algorithm called C3E (Consensus between Classification and Clustering Ensembles) was adapted and extended for the semi-supervised setting. This algorithm refines the sentiment classification from additional information provided by clusters of data, in a self-training procedure. This approach shows promising results when compared with state of the art algorithms.

Recommender systems (RSs) provide personalised suggestions of information or products relevant to user's needs. RSs are considered as powerful tools that help users to find interesting items matching their own taste. Although RSs have made substantial progress in theory and algorithm development and have achieved many commercial successes, how to utilise the widely available information on Online Social Networks (OSNs) has largely been overlooked. Noticing this gap in existing research on RSs and taking into account a user's selection being greatly influenced by his/her trusted friends and their opinions, this thesis proposes a novel personalised Recommender System framework, so-called Implicit Social Trust and Sentiment (ISTS) based RSs. The main motivation was to overcome the overlooked use of OSNs in Recommender Systems and to utilise the widely available information from such networks. This work also designs solutions to a number of challenges inherent to the RSs domain, such as accuracy, cold-start, diversity and coverage. ISTS improves the existing recommendation approaches by exploring a new source of data from friends' short posts in microbloggings. In the case of new users who have no previous preferences, ISTS maps the suggested recommendations into numerical rating scales by applying the three main components. The first component is measuring the implicit trust between friends based on their intercommunication activities and behaviour. Owing to the need to adapt friends' opinions, the implicit social trust model is designed to include the trusted friends and give them the highest weight of contribution in recommendation encounter. The second component is inferring the sentiment rating to reflect the knowledge behind friends' short posts, so-called micro-reviews. The sentiment behind micro-reviews is extracted using Sentiment Analysis (SA) techniques. To achieve the best sentiment representation, our approach considers the special natural environment in OSNs brief posts. Two Sentiment Analysis methodologies are used: a bag of words method and a probabilistic method. The third ISTS component is identifying the impact degree of friends' sentiments and their level of trust by using machine learning algorithms. Two types of machine learning algorithms are used: classification models and regressions models. The classification models include Naive Bayes, Logistic Regression and Decision Trees. Among the three classification models, Decision Trees show the best Mean absolute error (MAE) at 0.836. Support Vector Regression performed the best among all models at 0.45 of MAE. This thesis also proposes an approach with further improvement over ISTS, namely Hybrid Implicit Social Trust and Sentiment (H-ISTS). The enhanced approach applies improvements by optimising trust parameters to identify the impact of the features (re-tweets and followings/followers list) on recommendation results. Unlike the ISTS which allocates equal weight to trust features, H-ISTS provides different weights to determine the different effects of the two trust features. As a result, we found that H-ISTS improved the MAE to be 0.42 which is based on Support Vector Regression. Further, it increases the number of trust features from two to five features in order to include the influence of these features in rating predictions. The integration of the new approach H-ISTS with a Collaborative Filtering recommender system, in particular memory-based, is investigated next. Therefore, existing users with a history of ratings can receive recommendations by fusing their own tastes and their friends' preferences using the two type of memory-based methods: user-based and item-based. H-ISTSitem is the integration of H-ISTS and item-based which provides the lowest error at 0.7091. The experiments show that diversity is better achieved using the H-ISTSuser which is the integration of H-ISTS and user-based technique. To evaluate the performance of these approaches, two real social datasets are collected from Twitter. To verify the proposed framework, the experiments are conducted and the results are compared against the most relevant baselines which confirmed that RSs have been successfully improved using OSNs. These enhancements demonstrate the effectiveness and promises of the proposed approach in RSs.

Thèse numérisée par la Direction des bibliothèques de l'Université de Montréal.

With proliferation of user-generated reviews, new opportunities and challenges arise. The advance of Web technologies allows people to access a large amount of reviews of products and services online. Knowing what others like and dislike becomes increasingly important for their decision making in online shopping. The retailers also care more than ever about online reviews, because a vast pool of reviews enables them to monitor reputations and collect feedbacks efficiently. However, people often find difficult times in identifying and summarizing fine-grained sentiments buried in the opinion-rich resources. The traditional sentiment analysis, which focuses on the overall sentiments, fails to uncover the sentiments with regard to the aspects of the reviewed entities. This dissertation studied the research problem of Aspect Based Sentiment Analysis (ABSA), which is to reveal the aspect-dependent sentiment information of review text. ABSA consists of several subtasks: 1) aspect extraction, 2) aspect term extraction, 3) aspect category classification, and 4) sentiment polarity classification at aspect level. We focused on the approach of topic models and neural networks for ABSA. First, to extract the aspects from a collection of reviews and to detect the sentiment polarity regarding the aspects in each review, we proposed a few probabilistic graphical models, which can model words distribution in reviews and aspect ratings at the same time. Second, we presented a multi-task learning model based on long-short term memory and convolutional neural network for aspect category classification and aspect term extraction. Third, for aspect-level sentiment polarity classification, we developed a gated convolution neural network, which can be applied to aspect category sentiment analysis as well as aspect target sentiment analysis.

Les deux parties de ce mémoire révèlent -chacune à sa manière -mon sentiment d'attachement à la terre, au Nord, à la grande Amérique en même temps que le besoin viscéral de m'en affranchir parfois, de m'en éloigner pour mieux voir ce que chaque lieu imprime en moi à tout moment. La première partie, mon congé de l'amérique, est construite sous la forme d'une suite poétique où un moi-sujet se livre à un travail de décantation de l'affectif, du territorial et de l'identitaire. Chacune des cinq sections de ce long poème trace des portraits de l'Amérique qui originent aussi bien des paysages du Lac-Saint-Pierre, du Lac-Saint-Jean, des forêts du Nord, de Montréal que du monde intérieur présent dans chaque individu. L'exploration de ces territoires se fait par le biais du questionnement, de l'admiration, de la nostalgie, du deuil. L'Amérique se montre alors dans sa vastitude, ses névroses, ses beautés, ses contradictions, sa désuétude. Le moi-sujet, pour sa part, traverse ces aspérités, apprenant (parfois douloureusement) à composer avec elles. La seconde partie, Juste là, propose une réflexion formée d'un ensemble de rubriques portant sur les histoires que chaque lieu raconte à celui qui le traverse. Que ce soit par les bruits de la rue, les lignes tortueuses sur l'écorce d'un arbre, la douceur de l'herbe, le monde du dehors arrive certains jours à toucher des pans de l'être de façon inattendue. Ce sont des détails d'habitude, des éléments d'un décor familier qui, au lieu de passer inaperçus, tout à coup tendent la main pour montrer quelque chose et dire: «c'était juste là sous tes yeux». Alors, pendant un court laps de temps, l'individu se reconnaît dans ce qui se trouve devant lui et y participe intensément, profondément. Cette composition du dedans et du dehors, cette construction de l'être dans un imaginaire de la langue et du lieu, c'est le sens du travail d'atelier. L'atelier d'écriture, tout compte fait, permet de se situer, de s'ancrer dans un espace donné. Il faut arriver à s'y inscrire en tant que sujet parfois en s'altérant, parfois en s'imposant. Partout dans ce mémoire on pourra reconnaître les mêmes questions fondamentales : comment exister vraiment dans un lieu de grandeurs, dans un espace rempli d'histoires déjà, dans un monde qui ne correspond pas toujours à nos attentes et nos désirs? Comment les lieux déterminent-ils ce que l'on est, ce que l'on devient? Est-il possible enfin que moi aussi je puisse marquer les lieux que je traverse? ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : Amérique, Identité, Arbre, Appartenance, Nord, Désir, Deuil.

Ce mémoire en création littéraire comporte deux parties : un recueil de nouvelles, qui s'intitule Les bonnes filles plantent des fleurs au printemps, et un dossier d'accompagnement, que j'ai titré La seconde chance de l’écriture. Composée de douze nouvelles, la partie création met en place l'univers intime de femmes à différents épisodes de leur vie. Dans des périodes charnières marquées par la perte ou la peur de la perte (décès, deuils amoureux, maladie, conscience du vieillissement, nostalgie du passé), les protagonistes s'adonnent à la réflexion, à l'introspection. La presque totalité des textes sont écrits à la première personne, ce qui favorise une proximité, de sorte qu'on peut suivre les pensées, les impressions et les sentiments des narratrices, qui tentent d'exercer un certain contrôle sur ce qui leur échappe et cherchent un apaisement. Dans leur désarroi, elles remettent en question leurs convictions naïves sur le monde et sur les êtres qui les entourent, elles constatent que l'ordre des choses établies s'effondre, elles doivent abandonner leurs illusions et leur désir de perfection. Le style dépouillé, le langage mimant l'oralité et la brièveté de ces récits concentrés sur la trajectoire intérieure des personnages ancrent ces nouvelles dans une écriture de l'intime basée sur des rapports interpersonnels familiaux et sentimentaux problématiques. Le ton direct, souvent impudique, ainsi que l'importance d'un quotidien où il n'advient rien d'important, laisse davantage place à la confession qu'à l'intrigue. Ces textes appartiennent à ce qu'on appelle la nouvelle-instant. Le dossier d'accompagnement explore différents aspects d'un processus d'écriture basé sur l'expérience de la dépression. À partir de témoignages d'écrivaines ayant traversé des périodes difficiles, qui leur ont souvent été fatales (Virginia Woolf, Sylvia Plath), cet essai amorce une réflexion sur l'incidence des états dépressifs sur l'écriture. Souvent nécessaire à l'éveil de la création, la dépression apporte une lucidité que l'individu n'avait pas auparavant, laissant émerger une impudeur qui tient de la « franchise enfantine », selon le terme de Virginia Woolf. Quand, une fois rétabli, l'écrivain recommencera à écrire, il fera face à un dédoublement de la voix créatrice, celle de l'adulte et celle de l'enfant sans « surmoi ». La voix narrative surgira de la tension entre ces deux voix, qui doivent coexister afin que le texte trouve sa forme. L'écrivain est un funambule qui marche sur un fil tendu. Tout en reconnaissant le besoin de porter un regard distancié sur ce qu'il a vécu, il ne doit pas céder à l'autocensure. Le regard d'autrui et la peur d'être jugé l'obligent à livrer un combat constant, sans quoi sa pensée, dans ce qu'elle a de plus authentique, pourrait ne pas se déployer. L'œuvre d'autres créateurs (écrivains, artistes visuels, compositeurs, cinéastes) lui sont d'un grand secours pour ranimer l'impulsion créatrice et faire échec à la tentation de voiler sa pensée. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : écriture, nouvelles, autocensure, dépression, sentiments, enfance, impudeur.

This thesis is an investment strategy that seeks to profit from increases in market volatility. There have been several boom and bust cycles during the past fifteen years and volatility is projected to continue forward as a result of global asset misallocation and challenges stemming from debt liquidity. Volatility is measured by the Chicago Board of Options Exchange VIX volatility index. A proposed mean reversion strategy uses the VIX as a contrarian indicator of hope and fear to time decisions at extreme levels that have been determined through statistical analysis. This thesis found through back testing that market timing is possible at extreme levels of fear but is less reliable during extreme levels of hope and complacency. This strategy that utilizes measures of sentiment does however outperform the general market despite being active only five months on average per year. By synthesizing a broad range of fundamental, technical, and behavioral research, this thesis develops a unique contribution and practical set of market trading guidelines. The significance of these findings will help the individual investor to make better decisions during times of increased volatility.

More and more events and anomalies that have happened in recent years cannot be explained by traditional models, which leads to a pervasive doubt of the effectiveness of the efficient market hypothesis. In particular, over ninety percent of Taiwan¡¦s stock market investors are individuals, and the noise trading phenomenon is very common and has a great effect upon the return of stock. Hence, the measure of investor sentiment formed by noise traders becomes a task for the researcher studying the factors which effect the stock return in Taiwan. The objective of this paper is to find the investor sentiment proxy variables which can be a significant factor in explaining stock return. This analysis adopts the arbitrage pricing model of the macroeconomic factors. The sample contains data for most listed stocks on the Taiwan Stock Exchange from 1984 to 2002. By combining the stock or company characteristic related to the noise traders¡¦ perception, including market value, stock and etc., and phenomenons effect by investor sentiment, including closed-end fund discount, initial returns on IPOs, and number of IPOs to the arbitrage pricing model , we found that closed-end fund discount and initial returns on IPOs are significant and appropriate to investor sentiment proxy variables. However, the number of IPOs is not significant enough

The success of on-line review websites has led to an overwhelming number of on-line consumer reviews. These reviews have become an important tool for consumers when making a decision to purchase a product. This growth has led to the need for applications that enable this information to be presented in a way that is meaningful. These applications often rely on domain specific semantic lexicons which are both expensive and time consuming to make. The following thesis proposes an unsupervised approach for product aspect discovery in on-line consumer reviews. We apply a two step hierarchical clustering process in which we first cluster based on the semantic similarity of the contexts of terms and then on the similarity of the hypernyms of the cluster members. The method also includes a process for assigning class labels to each of the clusters. Finally an experiment showing how the proposed methods can be used to measure aspect based sentiment is performed. The methods proposed in this thesis are evaluated on a set of 157,865 reviews from a major commercial website and found that the two-step clustering process increases cluster F-scores over a single round of clustering. Finally, the proposed methods are compared to a state of the art topic modelling approach by Titov and McDonald (2008).