The purpose of this paper is to present an alternative perspective of marketing, which is lookingat the concept from a macro point of view, with a chosen country-Vietnam. There are three maingoals that this investigation strives to achieve. In specific, it aims first to observe whether theVietnamese consumers have favorable or unfavorable attitudes towards the marketing systemthat currently operating nationally. This is reached through an application of ICSM model, whichbased very much on the famous marketing mix paradigm (4Ps).The second purpose of this thesis is to test whether Vietnam market exist the relationshipsbetween consumers’ sentiment toward marketing practices and customer satisfaction. Theapplied models in this case are Kano Model and ACSI model where it proposed that therelationship exits. Thirdly, as Vietnam is a developing country, the role of government isexpected to have an effect on the issue of marketing practice, which implicitly can influence thevariables of consumer attitude and consumer satisfaction. Therefore, we took this opportunity todetermine whether it is the case and reflects that with results from previous studies.And above all, demographic factors including age, gender, education background, andoccupation are tested against the three main variables of attitude, satisfaction, and government.To achieve all the three main goals, we adopt quantitative research strategy with self-completedquestionnaires. A total sample of 445 respondents is obtained from two biggest cities ofVietnam- Hanoi and Ho Chi Minh City. The analyzed results show that Vietnamese consumersare much more favorable to the current marketing activities, in comparison to other countrieswith well-developed economic situation. Moreover, the results also proved that there is positiverelationship between the two pairs of variable: attitude - satisfaction, and government regulations- satisfaction; while between government and attitude, a negative correlation is found. Finally,the statistics also shows that demographic factors do have correlation on the variables thoughsome are more influential than others.

With the upsurge of Web 2.0, customers are able to share their opinions and feelings about products and services, politics, economic shifts, current events and any number of other topics on the Web. This information, if leveraged effectively, can provide rich and valuable insights, such as: input for vendors to create successful marketing strategies, understanding of areas of improvement in products and services and tracking political opinion. The problem with this information is that it is unorganised and unstructured, therefore, it is difficult to assess automatically and in bulk. Studies in the field of sentiment analysis aim to provide a solution to determining the polarities of, and gain an overview of, the wider public opinion behind certain topics in a large volume of textual data. This research provides a novel framework and a solid, practical implementation of the proposed framework for fine-grained sentiment analysis. The framework supports mixed-opinion text and multiword expressions when analysing the sentiments expressed and the aspects that those sentiments relate to. This research uses datasets across two domains in the customer reviews area (phone products and hotel services) to evaluate the proposed framework for its reliability and validity. A sizeable performance improvement was noted whereby the proposed methodology yielded a result of 91.3% accuracy in sentiment classification, as compared to the baseline (SentiWordNet), which had a result of 71.0%. In addition, an accuracy of 92.5% was observed for the aspect analysis automatically generated across the two domains tested.

In this thesis, we classified user comments posted on online forums related to “Newborn Genome Sequencing” (NGS). User comments were annotated as irrelevant, positive, negative, or mixed by two annotators. The objective was to create a classification model that could predict the sentiment of each user comment with a high accuracy. To compare classifiers, a baseline classifier (Accuracy 52%) was created. We created a single classifier (called flat comment-level classifier with accuracy of 65.14%) to classify comments into irrelevant, positive, negative, or mixed. A more sophisticated classifier, named two-level comment classifier, consisting of two classifiers, was created (Accuracy 69.81%): - The first classifier that classified each comment into relevant or irrelevant ones. - The second classifier that classified each relevant comment (predicted by the first classifier) as positive, negative, or mixed. 18 extra features were generated to improve the accuracy of the flat classification compared to baseline classifier (from 52% to 65.14% for flat comment classification, and 69.48% to 69.81% for two-level comment classification). Attempts were made to enhance the result of the two-level comment classifier by using the discourse structure of each sentence in a comment. The accuracy achieved by this enhanced two-level classifier was 64.24%. Therefore, removing irrelevant EDUs did not improve the accuracy. To achieve the above-mentioned enhancement, all comments were segmented into their consisting elementary discourse units (EDUs). We removed irrelevant EDUs from the relevant comments before running the second classifier. Furthermore, we performed EDU-level classification by creating two classifiers: - A flat classifier: classified all EDUs into irrelevant, positive, negative, or neutral - A two-level EDU: classified EDUs, first, into relevant or irrelevant and then classified the relevant EDUs (predicted by the first classifier) into positive, negative, or neutral ones. The accuracy achieved for the flat EDU-level classifier was 81.84%. However, due to the highly imbalanced nature of the EDU dataset, the F-measure for positive, negative, and neutral class was very low. Under-sampling was performed to improve the F-measure for positive, negative, and neutral class. Another topic investigated was to know why forum users supported or rejected NGS. To extract the arguments, the comments were segmented into EDUs. Following segmenting, each EDU was annotated as relevant or irrelevant to NGS. Each relevant EDU was annotated as for or against NGS. Topic related EDUs were selected as well as two EDUs before and after the topic-related EDUs. Bigrams, trigrams, four-grams and five-grams were created from extracted EDUs. Five-grams were more meaningful for human annotators, and were therefore favoured and ranked based on frequency in the dataset. Following ranking of the five grams, the top five were selected as the possible arguments.

This thesis studies the relation between the sentiment of individual investors and large-cap publicly traded corporations. The sentiment of individual investors is proxied by social media Facebook. In terms of investment decision making, the web or social media are main sources of information for individual investors. Due to technological development, the number of individual investors is growing. According to the behavioral finance theory, individual investors are considered as less rational agents that react to less rational information (emotions, social mood or opinion) irrelevant to fundamentals of securities. The impact of sentiment is studied at the level of volume of trade (liquidity) and stock prices. The results show a negative link between growing social media activity and volume of trade/stock prices. Delivered results can be considered as causal due to nature of employed data and presence of a weekend effect.

Indiana University-Purdue University Indianapolis (IUPUI) / Benefited from the booming social network, reading posts from other users over the internet is becoming one of commonest ways for people to intake information. One may also have noticed that sometimes we tend to focus on users provide well-founded analysis, rather than those merely who vent their emotions. This thesis aims at finding a simple and efficient way to recognize reliable information sources among countless internet users by examining the sentiments from their past posts. To achieve this goal, the research utilized a dataset of tweets about Apple's stock price retrieved from Twitter. Key features we studied include post-date, user name, number of followers of that user, and the sentiment of that tweet. Prior to making further use of the dataset, tweets from users who do not have sufficient posts are filtered out. To compare user sentiments and the derivative of Apple's stock price, we use Pearson correlation between them to describe how well each user performs. Then we iteratively increase the weight of reliable users and lower the weight of untrustworthy users, the correlation between overall sentiment and the derivative of stock price will finally converge. The final correlations for individual users are their performance scores. Due to the chaos of real-world data, manual segmentation via data visualization is also proposed as a denoise method to improve performance. Besides our method, other metrics can also be considered as user trust index, such as numbers of followers of each user. Experiments are conducted to prove that our method outperforms others. With simple input, this method can be applied to a wide range of topics including election, economy, and the job market.

This study aims to establish if a relationship between the investor sentiment generated from social media posts, such as Tweets, and the return on securities exists. If a relationship exists, one would be able to obtain an informational advantage from public information and outperform the market on a risk-adjusted basis. This would give the “outsider” information processed the predictive power of insider information, hence the title of the paper. The study makes use of Bloomberg's social activity data, which through natural language processing, allows for investor sentiment to be obtained by analysing a combination of Twitter and Stock Twits posts. This paper makes use of a three-prong approach, firstly examining if investor sentiment is a predictor of next-day returns. Next, an event study methodology is used to examine the optimal holding period, which can further be expanded to test market efficiency. Lastly, this paper considers the asymmetric risk aversion as outlined by Kahneman and Tversky (1979). Results show that there is little to no correlation between sentiment and next day returns. There is evidence for a multi-day holding period being optimal but statistically insignificant and there is no evidence found for asymmetric risk aversion.

Identifying differing opinions on a given topic as expressed by multiple people (as in a set of written reviews for a given product, for example) presents challenges. Opinions about a particular subject are often nuanced: a person may have both negative and positive opinions about different aspects of the subject of interest, and these aspect-specific opinions can be independent of the overall opinion on the subject. Being able to identify, collect, and count these nuanced opinions in a large set of data offers more insight into the strengths and weaknesses of competing products and services than does aggregating the overall ratings of such products and services. I make two useful and useable contributions in working with opinionated text. First, I present my implementation of a semi-supervised co-training machine classification method for identifying both product aspects (features of products) and sentiments expressed about such aspects. It offers better precision than fully-supervised methods while requiring much less text to be manually tagged (a time-consuming process). This algorithm can also be run in a fully supervised manner when more data is available. Second, I apply this co-training approach to reviews of restaurants and various electronic devices; such text contains both factual statements and opinions about features/aspects of products. The algorithm automatically identifies the product aspects and the words that indicate aspect-specific opinion polarity, while largely avoiding the problem of misclassifying the products themselves as inherently positive or negative. This method performs well compared to other approaches. When run on a set of reviews of five technology products collected from Amazon, the system performed with some demonstrated competence (with an average precision of 0.83) at the difficult task of simultaneously identifying aspects and sentiments, though comparison to contemporaries' simpler rules-based approaches was difficult. When run on a set of opinionated sentences about laptops and restaurants that formed the basis of a shared challenge in the SemEval-2014 Task 4 competition, it was able to classify the sentiments expressed about aspects of laptops better than any team that competed in the task (achieving 0.72 accuracy). It was above the mean in its ability to identify the aspects of restaurants about which people expressed opinions, even when co-training using only half of the labelled training data at the outset. While the SemEval-2014 aspect-based sentiment extraction task considered only separately the tasks of identifying product aspects and determining their polarities, I take an extra step and evaluate sentences as a whole, inferring aspects and the aspect-specific sentiments expressed simultaneously, a more difficult task that seems more applicable to real-world tasks. I present first results of this sentence-level task. The algorithm uses both lexical and syntactic information in a manner that is shown to be able to handle new words that it has never before seen. It offers some demonstrated ability to adapt to new subject domains for which it has no training data. The system is characterizable by very high precision and weak-to-average recall and it estimates its own confidence in its predictions; this characteristic should make the algorithm suitable for use on its own or for combination in a confidence-based voting ensemble. The software created for and described in the course of this dissertation is made available online.

Sentiment visualization techniques are information visualization approaches that focus on representing the results of sentiment analysis and opinion mining methods. Sentiment visualization techniques have been becoming more and more popular in the past few years, as demonstrated by recent surveys. Many techniques exist, and a lot of researchers and practitioners design their own. But the question of usability of these various techniques still remains generally unsolved, as the existing research typically addresses individual design alternatives for a particular technique implementation only. Multiple surveys and evaluations exist that argue for the importance of investigating the usability of such techniques further. This work focuses on evaluating the effectiveness, and efficiency of common visual representations for low-level visualization tasks in the context of sentiment visualization. It shows what previous work has already been done by other researchers and discusses the current state of the art. It further describes a task-based user study for various tasks, carried out as an online survey and taking the task completion time and error rate into account for most questions. This study is used for evaluating sentiment visualization techniques on their usability with regard to several sentiment and emotion datasets. This study shows that each visual representation and visual variable has its own weaknesses and strengths with respect to different tasks, which can be used as guidelines for future work in this area.

La gestion de classe est une fonction déterminante de l’efficacité de l’enseignement et de la qualité du climat d’apprentissage. Desbiens, Spallanzani, Roy, Turcotte, Lanoue et Tourigny (2011a) ont montré que l’établissement d’un climat favorable à l’apprentissage constitue un immense défi pour des stagiaires finissants en éducation physique et à la santé (ÉPS). Clairement, l’expérience du stage peut orienter la perception du stagiaire quant à sa capacité à remplir avec succès sa tâche d’enseignant. Selon la façon dont cette expérience est vécue, elle constituerait donc une occasion unique de développer leur sentiment d’efficacité personnelle (SEP) (Androzzi, 2011; Ballinger et Bishop, 2011). Cette recherche visait à répondre à quatre objectifs portant sur le SEP de stagiaires en enseignement de l’ÉPS. Plus spécifiquement elle cherchait à répondre à: a. mesurer l’écart de perception du SEP à gérer la classe entre deux moments, avant et après un stage; b. établir si les caractéristiques personnelles des stagiaires telles le sexe et l’âge influent de manière significative sur leur perception du SEP à gérer la classe avant et après le stage; c. établir si le degré d’expertise perçu dans les activités à enseigner durant le stage est modifié par l’expérience du stage en enseignement; d. établir si le degré d’expertise perçu dans les activités à enseigner durant le stage est relié au SEP à gérer la classe. Cette étude consiste en une analyse secondaire des données obtenues dans le cadre de l’étude réalisée par Desbiens et Spallanzani (2005). Elle a été menée auprès de 82 stagiaires de deux cohortes A (2003) et B (2004) d’un programme de formation initiale à l’enseignement de l’ÉPS au Québec (Canada) d’une durée de quatre ans. Ces stagiaires de deuxième année étaient inscrits à un cours de gestion de classe (année 2) ainsi qu’à un stage intensif de trois semaines spécifiquement centré sur la gestion de classe. Leurs perceptions quant au SEP à gérer la classe ont été mesurées à l’aide d’un questionnaire à choix forcés dans lequel les stagiaires devaient se prononcer par rapport à neuf composantes de la compétence à gérer la classe (MEQ, 2001). Les données collectées ont été traitées à l’aide de la version 13 du logiciel SPSS. Premièrement, les résultats démontrent qu’il y a six résultats significatifs pour la cohorte A dont trois ajustements positifs et trois ajustements négatifs et trois résultats significatifs pour la cohorte B qui consistent en des ajustements négatifs illustrant un ajustement à la baisse du SEP entre la fin et le début du stage. Deuxièmement, il semble que le stage n’apporte que quelques ajustements positifs du SEP au profit des stagiaires plus âgés. Troisièmement, quant aux différences de SEP entre les stagiaires des deux sexes, les résultats étaient différents entre la cohorte A (un ajustement à la hausse en faveur des stagiaires de sexe masculin) et la cohorte B (un ajustement à la hausse en faveur des stagiaires de sexe féminin). Quatrièmement, les résultats montrent que l’expérience du stage en enseignement n’apporte pas systématiquement et significativement un ajustement positif sur le degré d’expertise perçu dans les activités à enseigner. Finalement, nous constatons que la perception de maîtriser l’activité enseignée influence positivement et significativement la manière dont les stagiaires perçoivent leur SEP. Ces résultats convergent avec l’idée que le stage peut être une expérience marquante (Androzzi, 2011; Nichols, 2011), mais ils apportent des nuances à l’effet qu’il puisse être une source de développement du SEP (Bandura, 2007). De prochaines études mobilisant à la fois des approches qualitatives et quantitatives pourraient permettre de préciser à quelles conditions le stage peut contribuer au développement du SEP. L’étude des variables reliées au programme de formation et qui sont susceptibles d’influencer le développement du SEP des stagiaires sera une intéressante piste de recherche. D'autres recherches qui utilisent le construit d'efficacité personnelle pourraient également être menées auprès des enseignants associés qui reçoivent le stagiaire et qui l’accompagnent dans leur stage afin d’examiner leurs pratiques d’accompagnement et leurs effets sur le SEP. Ainsi, des études se penchant sur ces sujets pourraient s'avérer des avenues fortes intéressantes à explorer pour améliorer la formation initiale des stagiaires.

No description available.