Sentiment analysis on twitter for the portuguese language

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Business intelligence in banking: A literature analysis from 2002 to 2013 using Text Mining and latent Dirichlet allocation

telligence applications for the banking industry. Searches were performed in relevant journals resulting in 219 articles published between 2002 and 2013. To analyze such a large number of manuscripts, text mining techniques were used in pursuit for relevant terms on both business intelligence and banking domains. Moreover, the latent Dirichlet allocation modeling was used in or- der to group articles in several relevant topics. The analysis was conducted using a dictionary of terms belonging to both banking and business intelli- gence domains. Such procedure allowed for the identification of relationships between terms and topics grouping articles, enabling to emerge hypotheses regarding research directions. To confirm such hypotheses, relevant articles were collected and scrutinized, allowing to validate the text mining proce- dure. The results show that credit in banking is clearly the main application trend, particularly predicting risk and thus supporting credit approval or de- nial. There is also a relevant interest in bankruptcy and fraud prediction. Customer retention seems to be associated, although weakly, with targeting, justifying bank offers to reduce churn. In addition, a large number of ar- ticles focused more on business intelligence techniques and its applications, using the banking industry just for evaluation, thus, not clearly acclaiming for benefits in the banking business. By identifying these current research topics, this study also highlights opportunities for future research.

Publication trends of shared decision making in 15 high impact medical journals: a full-text review with bibliometric analysis.

BACKGROUND: Shared Decision Making (SDM) is increasingly advocated as a model for medical decision making. However, there is still low use of SDM in clinical practice. High impact factor journals might represent an efficient way for its dissemination. We aimed to identify and characterize publication trends of SDM in 15 high impact medical journals. METHODS: We selected the 15 general and internal medicine journals with the highest impact factor publishing original articles, letters and editorials. We retrieved publications from 1996 to 2011 through the full-text search function on each journal website and abstracted bibliometric data. We included publications of any type containing the phrase "shared decision making" or five other variants in their abstract or full text. These were referred to as SDM publications. A polynomial Poisson regression model with logarithmic link function was used to assess the evolution across the period of the number of SDM publications according to publication characteristics. RESULTS: We identified 1285 SDM publications out of 229,179 publications in 15 journals from 1996 to 2011. The absolute number of SDM publications by journal ranged from 2 to 273 over 16 years. SDM publications increased both in absolute and relative numbers per year, from 46 (0.32% relative to all publications from the 15 journals) in 1996 to 165 (1.17%) in 2011. This growth was exponential (P < 0.01). We found fewer research publications (465, 36.2% of all SDM publications) than non-research publications, which included non-systematic reviews, letters, and editorials. The increase of research publications across time was linear. Full-text search retrieved ten times more SDM publications than a similar PubMed search (1285 vs. 119 respectively). CONCLUSION: This review in full-text showed that SDM publications increased exponentially in major medical journals from 1996 to 2011. This growth might reflect an increased dissemination of the SDM concept to the medical community.

Publication trends of shared decision making in 15 high impact medical journals: a full-text review with bibliometric analysis.

BACKGROUND: Shared Decision Making (SDM) is increasingly advocated as a model for medical decision making. However, there is still low use of SDM in clinical practice. High impact factor journals might represent an efficient way for its dissemination. We aimed to identify and characterize publication trends of SDM in 15 high impact medical journals. METHODS: We selected the 15 general and internal medicine journals with the highest impact factor publishing original articles, letters and editorials. We retrieved publications from 1996 to 2011 through the full-text search function on each journal website and abstracted bibliometric data. We included publications of any type containing the phrase "shared decision making" or five other variants in their abstract or full text. These were referred to as SDM publications. A polynomial Poisson regression model with logarithmic link function was used to assess the evolution across the period of the number of SDM publications according to publication characteristics. RESULTS: We identified 1285 SDM publications out of 229,179 publications in 15 journals from 1996 to 2011. The absolute number of SDM publications by journal ranged from 2 to 273 over 16 years. SDM publications increased both in absolute and relative numbers per year, from 46 (0.32% relative to all publications from the 15 journals) in 1996 to 165 (1.17%) in 2011. This growth was exponential (P < 0.01). We found fewer research publications (465, 36.2% of all SDM publications) than non-research publications, which included non-systematic reviews, letters, and editorials. The increase of research publications across time was linear. Full-text search retrieved ten times more SDM publications than a similar PubMed search (1285 vs. 119 respectively). CONCLUSION: This review in full-text showed that SDM publications increased exponentially in major medical journals from 1996 to 2011. This growth might reflect an increased dissemination of the SDM concept to the medical community.

Source and Target Text-Orientation in Drama Translation. An Analysis of Four Translations of William Shakespeare’s Julius Caesar

Tässä pro gradu -tutkielmassa käsittelen lähde- ja kohdetekstikeskeisyyttä näytelmäkääntämisessä. Tutkimuskohteina olivat käännösten sanasto, syntaksi, näyttämötekniikka, kielikuvat, sanaleikit, runomitta ja tyyli. Tutkimuksen tarkoituksena oli selvittää, näkyykö teoreettinen painopisteen siirtyminen lähdetekstikeskeisyydestä kohdetekstikeskeisyyteen suomenkielisessä näytelmäkääntämisessä. Oletuksena oli, että siirtyminen näkyy käytetyissä käännösstrategioissa. Tutkimuksen teoriaosuudessa käsitellään ensin lähde- ja kohdetekstikeskeisiä käännösteorioita. Ensin esitellään kaksi lähdetekstikeskeistä teoriaa, jotka ovat Catfordin (1965) muodollinen vastaavuus ja Nidan (1964) dynaaminen ekvivalenssi. Kohdetekstikeskeisistä teorioista käsitellään Touryn (1980) ja Newmarkin (1981) teoreettisia näkemyksiä sekä Reiss ja Vermeerin (1986) esittelemää skopos-teoriaa. Vieraannuttamisen ja kotouttamisen periaatteet esitellään lyhyesti. Teoriaosuudessa käsitellään myös näytelmäkääntämistä, William Shakespearen kieltä ja siihen liittyviä käännösongelmia. Lisäksi esittelen lyhyesti Shakespearen kääntämistä Suomessa ja Julius Caesarin neljä kääntäjää. Tutkimuksen materiaalina oli neljä Shakespearen Julius Caesar –näytelmän suomennosta, joista Paavo Cajanderin käännös on julkaistu vuonna 1883, Eeva-Liisa Mannerin vuonna 1983, Lauri Siparin vuonna 2006 ja Jarkko Laineen vuonna 2007. Analyysissa käännöksiä verrattiin lähdetekstiin ja toisiinsa ja vertailtiin kääntäjien tekemiä käännösratkaisuja. Tulokset olivat oletuksen mukaisia. Lähdetekstikeskeisiä käännösstrategioita oli käytetty uusissa käännöksissä vähemmän kuin vanhemmissa. Kohdetekstikeskeiset strategiat erosivat huomattavasti toisistaan ja uusinta käännöstä voi sanoa adaptaatioksi. Jatkotutkimuksissa tulisi materiaali laajentaa koskemaan muitakin Shakespearen näytelmien suomennoksia. Eri aikakausien käännöksiä tulisi verrata keskenään ja toisiinsa, jotta voitaisiin luotettavasti kuvata muutosta lähde- ja kohdetekstikeskeisten käännösstrategioiden käytössä ja eri aikakausien tyypillisten strategioiden kartoittamiseksi.

Identification of research trends in the field of separation processes. Application of epidemiological model, citation analysis, text mining, and technical analysis of the financial markets

Choice of industrial development options and the relevant allocation of the research funds become more and more difficult because of the increasing R&D costs and pressure for shorter development period. Forecast of the research progress is based on the analysis of the publications activity in the field of interest as well as on the dynamics of its change. Moreover, allocation of funds is hindered by exponential growth in the number of publications and patents. Thematic clusters become more and more difficult to identify, and their evolution hard to follow. The existing approaches of research field structuring and identification of its development are very limited. They do not identify the thematic clusters with adequate precision while the identified trends are often ambiguous. Therefore, there is a clear need to develop methods and tools, which are able to identify developing fields of research. The main objective of this Thesis is to develop tools and methods helping in the identification of the promising research topics in the field of separation processes. Two structuring methods as well as three approaches for identification of the development trends have been proposed. The proposed methods have been applied to the analysis of the research on distillation and filtration. The results show that the developed methods are universal and could be used to study of the various fields of research. The identified thematic clusters and the forecasted trends of their development have been confirmed in almost all tested cases. It proves the universality of the proposed methods. The results allow for identification of the fast-growing scientific fields as well as the topics characterized by stagnant or diminishing research activity.

Target audience analysis in cyber domain

This thesis aims to find an effective way of conducting a target audience analysis (TAA) in cyber domain. There are two main focal points that are addressed; the nature of the cyber domain and the method of the TAA. Of the cyber domain the object is to find the opportunities, restrictions and caveats that result from its digital and temporal nature. This is the environment in which the TAA method is examined in this study. As the TAA is an important step of any psychological operation and critical to its success, the method used must cover all the main aspects affecting the choice of a proper target audience. The first part of the research was done by sending an open-ended questionnaire to operators in the field of information warfare both in Finland and abroad. As the results were inconclusive, the research was completed by assessing the applicability of United States Army Joint Publication FM 3-05.301 in the cyber domain via a theory-based content analysis. FM 3- 05.301 was chosen because it presents a complete method of the TAA process. The findings were tested against the results of the questionnaire and new scientific research in the field of psychology. The cyber domain was found to be “fast and vast”, volatile and uncontrollable. Although governed by laws to some extent, the cyber domain is unpredictable by nature and not controllable to reasonable amount. The anonymity and lack of verification often present in the digital channels mean that anyone can have an opinion, and any message sent may change or even be counterproductive to the original purpose. The TAA method of the FM 3-05.301 is applicable in the cyber domain, although some parts of the method are outdated and thus suggested to be updated if used in that environment. The target audience categories of step two of the process were replaced by new groups that exist in the digital environment. The accessibility assessment (step eight) was also redefined, as in the digital media the mere existence of a written text is typically not enough to convey the intended message to the target audience. The scientific studies made in computer sciences and both in psychology and sociology about the behavior of people in social media (and overall in cyber domain) call for a more extensive remake of the TAA process. This falls, however, out of the scope of this work. It is thus suggested that further research should be carried out in search of computer-assisted methods and a more thorough TAA process, utilizing the latest discoveries of human behavior. ---------------------------------------------------------------------------------------------------------------------------------- Tämän opinnäytetyön tavoitteena on löytää tehokas tapa kohdeyleisöanalyysin tekemiseksi kybertoimintaympäristössä. Työssä keskitytään kahteen ilmiöön: kybertoimintaympäristön luonteeseen ja kohdeyleisöanalyysin metodiin. Kybertoimintaympäristön osalta tavoitteena on löytää sen digitaalisesta ja ajallisesta luonteesta juontuvat mahdollisuudet, rajoitteet ja sudenkuopat. Tämä on se ympäristö jossa kohdeyleisöanalyysiä tarkastellaan tässä työssä. Koska kohdeyleisöanalyysi kuuluu olennaisena osana jokaiseen psykologiseen operaatioon ja on onnistumisen kannalta kriittinen tekijä, käytettävän metodin tulee pitää sisällään kaikki oikean kohdeyleisön valinnan kannalta merkittävät osa-alueet. Tutkimuksen ensimmäisessä vaiheessa lähetettiin avoin kysely informaatiosodankäynnin ammattilaisille Suomessa ja ulkomailla. Koska kyselyn tulokset eivät olleet riittäviä johtopäätösten tekemiseksi, tutkimusta jatkettiin tarkastelemalla Yhdysvaltojen armeijan kenttäohjesäännön FM 3-05.301 soveltuvuutta kybertoimintaympäristössä käytettäväksi teorialähtöisen sisällönanalyysin avulla. FM 3-05.301 valittiin koska se sisältää kokonaisvaltaisen kohdeyleisöanalyysiprosessin. Havaintoja verrattiin kyselytutkimuksen tuloksiin ja psykologian uusiin tutkimuksiin. Kybertoimintaympäristö on tulosten perusteella nopea ja valtava, jatkuvasti muuttuva ja kontrolloimaton. Vaikkakin lait hallitsevat kybertoimintaympäristöä jossakin määrin, on se silti luonteeltaan ennakoimaton eikä sitä voida luotettavasti hallita. Digitaalisilla kanavilla usein läsnäoleva nimettömyys ja tiedon tarkastamisen mahdottomuus tarkoittavat että kenellä tahansa voi olla mielipide asioista, ja mikä tahansa viesti voi muuttua, jopa alkuperäiseen tarkoitukseen nähden vastakkaiseksi. FM 3-05.301:n metodi toimii kybertoimintaympäristössä, vaikkakin jotkin osa-alueet ovat vanhentuneita ja siksi ne esitetään päivitettäväksi mikäli metodia käytetään kyseisessä ympäristössä. Kohdan kaksi kohdeyleisökategoriat korvattiin uusilla, digitaalisessa ympäristössä esiintyvillä ryhmillä. Lähestyttävyyden arviointi (kohta 8) muotoiltiin myös uudestaan, koska digitaalisessa mediassa pelkkä tekstin läsnäolo ei sellaisenaan tyypillisesti vielä riitä halutun viestin välittämiseen kohdeyleisölle. Tietotekniikan edistyminen ja psykologian sekä sosiologian aloilla tehty tieteellinen tutkimus ihmisten käyttäytymisestä sosiaalisessa mediassa (ja yleensä kybertoimintaympäristössä) mahdollistavat koko kohdeyleisöanalyysiprosessin uudelleenrakentamisen. Tässä työssä sitä kuitenkaan ei voida tehdä. Siksi esitetäänkin että lisätutkimusta tulisi tehdä sekä tietokoneavusteisten prosessien että vielä syvällisempien kohdeyleisöanalyysien osalta, käyttäen hyväksi viimeisimpiä ihmisen käyttäytymiseen liittyviä tutkimustuloksia.

Measuring Effectiveness of Marketing Message: The Role of Verbal, Visual and Brand Elements

The aim of this paper is to study the role of verbal, visual and brand elements while meas-uring effectiveness of marketing message. The thesis is written in the context of mobile gaming industry. The object of the study is marketing message. To achieve the aim, the main research question was formulated: How do the elements of marketing message, such as verbal, visual and brand, affect the consumer’s attitude toward the ad, emotional response and attention capture? The theory development chapter lays on three corner stones – analysis of previous litera-ture on marketing message and its elements, namely verbal, visual and brand; overview of literature on attitude formation and particularly attitude toward the ad. In addition, investiga-tion of key points of emotional response and attention capture literature finalizes the chap-ter. The empirical part consists of experiment, conducted with 27 participants. Experiment includes the self-report semantically anchored scale, measuring the attitude toward the ad, as well as autonomic measures – eye tracking (attention capture) and facial expressions (emotional response). The results of the experiment showed that the size of the brand element – the logo – has an effect on the attention capture and the overall attitude toward the ad. The bigger the logo, the more time people spend viewing it, and they realise the message is more educa-tional and factual. The measure related to the visual element – the visual complexity – in-creases the intensity of participant’s facial expression. While the measure of verbal ele-ment – the contrast between text and background colours – leads to a better attitude to-ward the ad. The higher the contrast between text and background, the more known the message appears to the viewer.

The relation between content and structure in language production: an analysis of speech errors in semantic dementia

In order to explore the impact of a degraded semantic system on the structure of language production, we analysed transcripts from autobiographical memory interviews to identify naturally-occurring speech errors by eight patients with semantic dementia (SD) and eight age-matched normal speakers. Relative to controls, patients were significantly more likely to (a) substitute and omit open class words, (b) substitute (but not omit) closed class words, (c) substitute incorrect complex morphological forms and (d) produce semantically and/or syntactically anomalous sentences. Phonological errors were scarce in both groups. The study confirms previous evidence of SD patients’ problems with open class content words which are replaced by higher frequency, less specific terms. It presents the first evidence that SD patients have problems with closed class items and make syntactic as well as semantic speech errors, although these grammatical abnormalities are mostly subtle rather than gross. The results can be explained by the semantic deficit which disrupts the representation of a pre-verbal message, lexical retrieval and the early stages of grammatical encoding.

The discourses on induced abortion in Ugandan daily newspapers : a discourse analysis

Background: Ugandan law prohibits abortion under all circumstances except where there is a risk for the woman's life. However, it has been estimated that over 250 000 illegal abortions are being performed in the country yearly. Many of these abortions are carried out under unsafe conditions, being one of the most common reasons behind the nearly 5000 maternal deaths per year in Uganda. Little research has been conducted in relation to societal views on abortion within the Ugandan society. This study aims to analyze the discourse on abortion as expressed in the two main daily Ugandan newspapers. Method: The conceptual content of 59 articles on abortion between years 2006-2012, from the two main daily English-speaking newspapers in Uganda, was studied using principles from critical discourse analysis. Results: A religious discourse and a human rights discourse, together with medical and legal sub discourses frame the subject of abortion in Uganda, with consequences for who is portrayed as a victim and who is to blame for abortions taking place. It shows the strong presence of the Catholic Church within the medial debate on abortion. The results also demonstrate the absence of medial statements related to abortion made by political stakeholders. Conclusions: The Catholic Church has a strong position within the Ugandan society and their stance on abortion tends to have great influence on the way other actors and their activities are presented within the media, as well as how stakeholders choose to convey their message, or choose not to publicly debate the issue in question at all. To decrease the number of maternal deaths, we highlight the need for a more inclusive and varied debate that problematizes the current situation, especially from a gender perspective.

Chemistry and Chemical Education Through Text and Image: Analysis of Twentieth Century Textbooks Used in Brazilian Context

Assuming that textbooks give literary expression to cultural and ideological values of a nation or group, we propose the analysis of chemistry textbooks used in Brazilian universities throughout the twentieth century. We analyzed iconographic and textual aspects of 31 textbooks which had significant diffusion in the context of Brazilian universities at that period. As a result of the iconographic analysis, nine categories of images were proposed: (1) laboratory and experimentation, (2) industry and production, (3) graphs and diagrams, (4) illustrations related to daily life, (5) models, (6) illustrations related to the history of science, (7) pictures or diagrams of animal, vegetable or mineral samples, (8) analogies and (9) concepts of physics. The distribution of images among the categories showed a different emphasis in the presentation of chemical content due to a commitment to different conceptions of chemistry over the period. So, we started with chemistry as an experimental science in the early twentieth century, with an emphasis change to the principles of chemistry from the 1950s, culminating in a chemistry of undeniable technological influence. Results showed that reflections not only on the history of science, but on the history of science education, may be useful for the improvement of science education.

Dynamic networks, text analysis and Gephi: the art math

In numerosi campi scientici l'analisi di network complessi ha portato molte recenti scoperte: in questa tesi abbiamo sperimentato questo approccio sul linguaggio umano, in particolare quello scritto, dove le parole non interagiscono in modo casuale. Abbiamo quindi inizialmente presentato misure capaci di estrapolare importanti strutture topologiche dai newtork linguistici(Degree, Strength, Entropia, . . .) ed esaminato il software usato per rappresentare e visualizzare i grafi (Gephi). In seguito abbiamo analizzato le differenti proprietà statistiche di uno stesso testo in varie sue forme (shuffolato, senza stopwords e senza parole con bassa frequenza): il nostro database contiene cinque libri di cinque autori vissuti nel XIX secolo. Abbiamo infine mostrato come certe misure siano importanti per distinguere un testo reale dalle sue versioni modificate e perché la distribuzione del Degree di un testo normale e di uno shuffolato abbiano lo stesso andamento. Questi risultati potranno essere utili nella sempre più attiva analisi di fenomeni linguistici come l'autorship attribution e il riconoscimento di testi shuffolati.