Learning domain-specific sentiment lexicons with applications to recommender systems

Search is now going beyond looking for factual information, and people wish to search for the opinions of others to help them in their own decision-making. Sentiment expressions or opinion expressions are used by users to express their opinion and embody important pieces of information, particularly in online commerce. The main problem that the present dissertation addresses is how to model text to find meaningful words that express a sentiment. In this context, I investigate the viability of automatically generating a sentiment lexicon for opinion retrieval and sentiment classification applications. For this research objective we propose to capture sentiment words that are derived from online users’ reviews. In this approach, we tackle a major challenge in sentiment analysis which is the detection of words that express subjective preference and domain-specific sentiment words such as jargon. To this aim we present a fully generative method that automatically learns a domain-specific lexicon and is fully independent of external sources. Sentiment lexicons can be applied in a broad set of applications, however popular recommendation algorithms have somehow been disconnected from sentiment analysis. Therefore, we present a study that explores the viability of applying sentiment analysis techniques to infer ratings in a recommendation algorithm. Furthermore, entities’ reputation is intrinsically associated with sentiment words that have a positive or negative relation with those entities. Hence, is provided a study that observes the viability of using a domain-specific lexicon to compute entities reputation. Finally, a recommendation system algorithm is improved with the use of sentiment-based ratings and entities reputation.

COMPENDIUM: a text summarisation tool for generating summaries of multiple purposes, domains, and genres

In this paper, we present a Text Summarisation tool, compendium, capable of generating the most common types of summaries. Regarding the input, single- and multi-document summaries can be produced; as the output, the summaries can be extractive or abstractive-oriented; and finally, concerning their purpose, the summaries can be generic, query-focused, or sentiment-based. The proposed architecture for compendium is divided in various stages, making a distinction between core and additional stages. The former constitute the backbone of the tool and are common for the generation of any type of summary, whereas the latter are used for enhancing the capabilities of the tool. The main contributions of compendium with respect to the state-of-the-art summarisation systems are that (i) it specifically deals with the problem of redundancy, by means of textual entailment; (ii) it combines statistical and cognitive-based techniques for determining relevant content; and (iii) it proposes an abstractive-oriented approach for facing the challenge of abstractive summarisation. The evaluation performed in different domains and textual genres, comprising traditional texts, as well as texts extracted from the Web 2.0, shows that compendium is very competitive and appropriate to be used as a tool for generating summaries.

New markov chain based methods for single and cross-domain sentiment classification

Nowadays communication is switching from a centralized scenario, where communication media like newspapers, radio, TV programs produce information and people are just consumers, to a completely different decentralized scenario, where everyone is potentially an information producer through the use of social networks, blogs, forums that allow a real-time worldwide information exchange. These new instruments, as a result of their widespread diffusion, have started playing an important socio-economic role. They are the most used communication media and, as a consequence, they constitute the main source of information enterprises, political parties and other organizations can rely on. Analyzing data stored in servers all over the world is feasible by means of Text Mining techniques like Sentiment Analysis, which aims to extract opinions from huge amount of unstructured texts. This could lead to determine, for instance, the user satisfaction degree about products, services, politicians and so on. In this context, this dissertation presents new Document Sentiment Classification methods based on the mathematical theory of Markov Chains. All these approaches bank on a Markov Chain based model, which is language independent and whose killing features are simplicity and generality, which make it interesting with respect to previous sophisticated techniques. Every discussed technique has been tested in both Single-Domain and Cross-Domain Sentiment Classification areas, comparing performance with those of other two previous works. The performed analysis shows that some of the examined algorithms produce results comparable with the best methods in literature, with reference to both single-domain and cross-domain tasks, in $2$-classes (i.e. positive and negative) Document Sentiment Classification. However, there is still room for improvement, because this work also shows the way to walk in order to enhance performance, that is, a good novel feature selection process would be enough to outperform the state of the art. Furthermore, since some of the proposed approaches show promising results in $2$-classes Single-Domain Sentiment Classification, another future work will regard validating these results also in tasks with more than $2$ classes.

Towards a unified sentiment lexicon based on graphics processing units

This paper presents an approach to create what we have called a Unified Sentiment Lexicon (USL). This approach aims at aligning, unifying, and expanding the set of sentiment lexicons which are available on the web in order to increase their robustness of coverage. One problem related to the task of the automatic unification of different scores of sentiment lexicons is that there are multiple lexical entries for which the classification of positive, negative, or neutral {P, Z, N} depends on the unit of measurement used in the annotation methodology of the source sentiment lexicon. Our USL approach computes the unified strength of polarity of each lexical entry based on the Pearson correlation coefficient which measures how correlated lexical entries are with a value between 1 and -1, where 1 indicates that the lexical entries are perfectly correlated, 0 indicates no correlation, and -1 means they are perfectly inversely correlated and so is the UnifiedMetrics procedure for CPU and GPU, respectively. Another problem is the high processing time required for computing all the lexical entries in the unification task. Thus, the USL approach computes a subset of lexical entries in each of the 1344 GPU cores and uses parallel processing in order to unify 155802 lexical entries. The results of the analysis conducted using the USL approach show that the USL has 95.430 lexical entries, out of which there are 35.201 considered to be positive, 22.029 negative, and 38.200 neutral. Finally, the runtime was 10 minutes for 95.430 lexical entries; this allows a reduction of the time computing for the UnifiedMetrics by 3 times.

Generating Linked-Data based Domain-Specific Sentiment Lexicons from Legacy Language and Semantic Resources

We present a methodology for legacy language resource adaptation that generates domain-specific sentiment lexicons organized around domain entities described with lexical information and sentiment words described in the context of these entities. We explain the steps of the methodology and we give a working example of our initial results. The resulting lexicons are modelled as Linked Data resources by use of established formats for Linguistic Linked Data (lemon, NIF) and for linked sentiment expressions (Marl), thereby contributing and linking to existing Language Resources in the Linguistic Linked Open Data cloud.

Linked-data based domain-specific sentiment lexicons

In this paper we present a dataset componsed of domain-specific sentiment lexicons in six languages for two domains. We used existing collections of reviews from Trip Advisor, Amazon, the Stanford Network Analysis Project and the OpinRank Review Dataset. We use an RDF model based on the lemon and Marl formats to represent the lexicons. We describe the methodology that we applied to generate the domain-specific lexicons and we provide access information to our datasets.

Development of a Dashboard for Sentiment Analysis of Football in Twitter based on Web Components and D3.js

This thesis is the result of a project whose objective has been to develop and deploy a dashboard for sentiment analysis of football in Twitter based on web components and D3.js. To do so, a visualisation server has been developed in order to present the data obtained from Twitter and analysed with Senpy. This visualisation server has been developed with Polymer web components and D3.js. Data mining has been done with a pipeline between Twitter, Senpy and ElasticSearch. Luigi have been used in this process because helps building complex pipelines of batch jobs, so it has analysed all tweets and stored them in ElasticSearch. To continue, D3.js has been used to create interactive widgets that make data easily accessible, this widgets will allow the user to interact with them and �filter the most interesting data for him. Polymer web components have been used to make this dashboard according to Google's material design and be able to show dynamic data in widgets. As a result, this project will allow an extensive analysis of the social network, pointing out the influence of players and teams and the emotions and sentiments that emerge in a lapse of time.

Text classification using compression-based dissimilarity measures

Arguably, the most difficult task in text classification is to choose an appropriate set of features that allows machine learning algorithms to provide accurate classification. Most state-of-the-art techniques for this task involve careful feature engineering and a pre-processing stage, which may be too expensive in the emerging context of massive collections of electronic texts. In this paper, we propose efficient methods for text classification based on information-theoretic dissimilarity measures, which are used to define dissimilarity-based representations. These methods dispense with any feature design or engineering, by mapping texts into a feature space using universal dissimilarity measures; in this space, classical classifiers (e.g. nearest neighbor or support vector machines) can then be used. The reported experimental evaluation of the proposed methods, on sentiment polarity analysis and authorship attribution problems, reveals that it approximates, sometimes even outperforms previous state-of-the-art techniques, despite being much simpler, in the sense that they do not require any text pre-processing or feature engineering.

Sentiment analysis on twitter for the portuguese language

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Els catalans de sentiment nacional espanyol i el seu suport a la sobirania : Condicions i Drets.

L'objecte d'aquest treball final del Màster d'Anàlisi Política de la UOC (2008) és fer una reflexió basada en casos de diversos països de característiques similars a Catalunya, seguida de propostes estratègiques i de decisions a prendre sobre les condicions i seguretats que podrien fer decantar la població catalana de sentiment nacional espanyol cap a un vot favorable a la independència en un referèndum sobiranista.

EMU sovereign debt market crisis: Fundamentals-based or pure contagion?

We empirically investigate whether the transmission of the recent crisis in euro area sovereign debt markets was due to fundamentals-based or pure contagion. To do so, we examine the behaviour of EMU sovereign bond yield spreads with respect to the German bund for a sample of both central and peripheral countries from January 1999 to December 2012. First we apply a dynamic approach to analyse the evolution of the degree of Grangercausality within the 90 pairs of sovereign bond yield spreads in our sample, in order to detect episodes of significantly increased causality between them (which we associate with contagion) and episodes of significantly reduced interconnection (which we associate with immunisation). We then use an ordered logit model to assess the determinants of the occurrence of the episodes detected. Our results suggest the importance of variables proxying market sentiment and of variables proxying macrofundamentals in determining contagion and immunisation outcomes. Therefore, our findings underline the coexistence of “pure” and “fundamentals-based contagion” during the recent European debt crisis.

EMU sovereign debt market crisis: Fundamentals-based or pure contagion?

We empirically investigate whether the transmission of the recent crisis in euro area sovereign debt markets was due to fundamentals-based or pure contagion. To do so, we examine the behaviour of EMU sovereign bond yield spreads with respect to the German bund for a sample of both central and peripheral countries from January 1999 to December 2012. First we apply a dynamic approach to analyse the evolution of the degree of Grangercausality within the 90 pairs of sovereign bond yield spreads in our sample, in order to detect episodes of significantly increased causality between them (which we associate with contagion) and episodes of significantly reduced interconnection (which we associate with immunisation). We then use an ordered logit model to assess the determinants of the occurrence of the episodes detected. Our results suggest the importance of variables proxying market sentiment and of variables proxying macrofundamentals in determining contagion and immunisation outcomes. Therefore, our findings underline the coexistence of “pure” and “fundamentals-based contagion” during the recent European debt crisis.

Le sentiment de gratitude en soins palliatifs : un facteur contribuant au bien-être et à la qualité de vie ? [Gratitude in palliative care: is it a psychological factor contributing to well-being and quality of life?]

En s'appuyant sur la littérature scientifique et l'état de la recherche, cet article a pour objectif de montrer pourquoi la prise en compte du sentiment de gratitude peut s'avérer utile dans le contexte palliatif en mettant en évidence dans quelle mesure cette émotion ou disposition individuelle: 1) entre en résonance avec le concept de croissance posttraumatique et certains enjeux relationnels chez les patients en fin de vie; 2) représente un facteur favorisant le bien-être et la qualité de vie; 3) peut être considérée comme un facteur protecteur contre les troubles psychopathologiques. Based on the scientific literature and the state of research, this article aims to show why the feeling of gratitude may represent a point of interest for palliative care. We will highlight the following in this article: 1) why this feeling of gratitude resonates with the concept of post-traumatic growth and relational challenges in end-of-life patients; 2) in which measure this feeling represents a factor contributing to well-being and quality of life; 3) in which measure this feeling may be considered as a protective factor against psychopathological troubles.

PERFORMANCE ANALYSIS OF A TACTICAL ASSET ALLOCATION – Evidence from the Black–Litterman model based fixed income portfolio

Over time the demand for quantitative portfolio management has increased among financial institutions but there is still a lack of practical tools. In 2008 EDHEC Risk and Asset Management Research Centre conducted a survey of European investment practices. It revealed that the majority of asset or fund management companies, pension funds and institutional investors do not use more sophisticated models to compensate the flaws of the Markowitz mean-variance portfolio optimization. Furthermore, tactical asset allocation managers employ a variety of methods to estimate return and risk of assets, but also need sophisticated portfolio management models to outperform their benchmarks. Recent development in portfolio management suggests that new innovations are slowly gaining ground, but still need to be studied carefully. This thesis tries to provide a practical tactical asset allocation (TAA) application to the Black–Litterman (B–L) approach and unbiased evaluation of B–L models’ qualities. Mean-variance framework, issues related to asset allocation decisions and return forecasting are examined carefully to uncover issues effecting active portfolio management. European fixed income data is employed in an empirical study that tries to reveal whether a B–L model based TAA portfolio is able outperform its strategic benchmark. The tactical asset allocation utilizes Vector Autoregressive (VAR) model to create return forecasts from lagged values of asset classes as well as economic variables. Sample data (31.12.1999–31.12.2012) is divided into two. In-sample data is used for calibrating a strategic portfolio and the out-of-sample period is for testing the tactical portfolio against the strategic benchmark. Results show that B–L model based tactical asset allocation outperforms the benchmark portfolio in terms of risk-adjusted return and mean excess return. The VAR-model is able to pick up the change in investor sentiment and the B–L model adjusts portfolio weights in a controlled manner. TAA portfolio shows promise especially in moderately shifting allocation to more risky assets while market is turning bullish, but without overweighting investments with high beta. Based on findings in thesis, Black–Litterman model offers a good platform for active asset managers to quantify their views on investments and implement their strategies. B–L model shows potential and offers interesting research avenues. However, success of tactical asset allocation is still highly dependent on the quality of input estimates.

L’incidence d’un dispositif de soutien en gestion de classe sur les pratiques disciplinaires et le sentiment d’efficacité d’enseignants débutants

Résumé Cette étude quasi expérimentale consistait à élaborer et à mettre à l’essai une mesure de soutien à l’intention d’enseignants débutants ainsi qu’à évaluer l’efficacité de celle-ci. L’une des particularités de cette mesure, appelée Dispositif de soutien en gestion de classe, était qu’elle était centrée essentiellement sur le développement de la compétence à gérer la classe. L’application du dispositif, échelonnée sur une année scolaire, portait sur une trentaine d’enseignants débutants œuvrant au primaire, en milieu défavorisé, à Montréal. Basé sur les trois phases du modèle théorique d’Archambault et Chouinard (2003), le dispositif se déclinait selon trois cycles de formation : l’établissement du fonctionnement de la classe, le maintien de celui-ci et le soutien à la motivation scolaire, ainsi que l’intervention pour résoudre des problèmes de comportement. Chaque cycle commençait par une journée de formation et d’appropriation (JFA) durant laquelle il y avait présentation d’un contenu théorique puis des ateliers d’appropriation. Par la suite, les enseignants effectuaient des mises en pratique dans leur classe. Pour terminer le cycle, un autre type de rencontre, la rencontre de suivi (RS), servait entre autres à objectiver la pratique. L’aspect original de cette mesure de soutien était que la première rencontre de formation était offerte une semaine avant la rentrée scolaire. Sur le thème « Commencer l’année du bon pied en gestion de classe », cette journée avait pour objectif de soutenir les enseignants débutants dans l’installation du fonctionnement de leur classe. L’efficacité du dispositif a été évaluée sur la base de trois dimensions : l’établissement et le maintien de l’ordre et de la discipline, le sentiment d’efficacité personnelle ainsi que la motivation professionnelle. Les perceptions du groupe d’enseignants débutants ayant pris part aux activités du dispositif (n = 27) ont été comparées à celles d’un groupe témoin (n = 44). Les participants avaient, en moyenne, 2,9 années d’expérience et leur âge variait de 23 à 56 ans. Les données ont été recueillies à l’aide d’un questionnaire auto rapporté rempli en deux temps, soit au deuxième et au huitième mois de l’année scolaire. Les scores des enseignants débutants du dispositif ont augmenté dans le temps pour l’ensemble des variables à l’étude. De plus, les analyses de variance à mesures répétées ont révélé que le dispositif a eu une triple incidence positive, attestée par des effets d’interaction. Les enseignants débutants engagés dans la démarche ont connu une augmentation de leur capacité à implanter les règles de classe, de leur sentiment d’efficacité personnelle à gérer les situations d’apprentissage et de leur motivation professionnelle. En effet, alors que, au début de l’étude, ils rapportaient des scores significativement inférieurs à ceux du groupe témoin, à la fin, les scores étaient équivalents. Les résultats ont aussi montré que les participants du groupe expérimental se distinguaient en affichant un meilleur sentiment d’efficacité à faire apprendre leurs élèves. L’étude nous apprend également que le sentiment d’efficacité personnelle à faire face aux problèmes de comportement et la capacité à gérer les comportements se sont renforcés de façon significative dans le temps chez l’ensemble des enseignants débutants. Finalement, aucun changement significatif n’a été détecté pour deux des huit variables à l’étude : le sentiment d’efficacité personnelle à avoir un effet sur le comportement des élèves et l’application des règles de classe. En définitive, ces résultats sont encourageants. Ils montrent l’enrichissement professionnel que les enseignants débutants peuvent retirer lorsqu’ils sont soutenus adéquatement. Nous croyons que la journée de formation portant sur l’installation du fonctionnement de la classe, avant la rentrée scolaire, a joué un rôle central dans les succès vécus par les enseignants débutants participants. C’est pourquoi nous recommandons ce type de formation assorti d’un suivi à long terme, où d’autres composantes entrent en jeu, afin de nourrir le sentiment d’efficacité personnelle et la motivation professionnelle des nouveaux enseignants.