A Curated Corpus for Sentiment-Topic Analysis

Postprint

Contextual text mining

With the dramatic growth of text information, there is an increasing need for powerful text mining systems that can automatically discover useful knowledge from text. Text is generally associated with all kinds of contextual information. Those contexts can be explicit, such as the time and the location where a blog article is written, and the author(s) of a biomedical publication, or implicit, such as the positive or negative sentiment that an author had when she wrote a product review; there may also be complex context such as the social network of the authors. Many applications require analysis of topic patterns over different contexts. For instance, analysis of search logs in the context of the user can reveal how we can improve the quality of a search engine by optimizing the search results according to particular users; analysis of customer reviews in the context of positive and negative sentiments can help the user summarize public opinions about a product; analysis of blogs or scientific publications in the context of a social network can facilitate discovery of more meaningful topical communities. Since context information significantly affects the choices of topics and language made by authors, in general, it is very important to incorporate it into analyzing and mining text data. In general, modeling the context in text, discovering contextual patterns of language units and topics from text, a general task which we refer to as Contextual Text Mining, has widespread applications in text mining. In this thesis, we provide a novel and systematic study of contextual text mining, which is a new paradigm of text mining treating context information as the ``first-class citizen.'' We formally define the problem of contextual text mining and its basic tasks, and propose a general framework for contextual text mining based on generative modeling of text. This conceptual framework provides general guidance on text mining problems with context information and can be instantiated into many real tasks, including the general problem of contextual topic analysis. We formally present a functional framework for contextual topic analysis, with a general contextual topic model and its various versions, which can effectively solve the text mining problems in a lot of real world applications. We further introduce general components of contextual topic analysis, by adding priors to contextual topic models to incorporate prior knowledge, regularizing contextual topic models with dependency structure of context, and postprocessing contextual patterns to extract refined patterns. The refinements on the general contextual topic model naturally lead to a variety of probabilistic models which incorporate different types of context and various assumptions and constraints. These special versions of the contextual topic model are proved effective in a variety of real applications involving topics and explicit contexts, implicit contexts, and complex contexts. We then introduce a postprocessing procedure for contextual patterns, by generating meaningful labels for multinomial context models. This method provides a general way to interpret text mining results for real users. By applying contextual text mining in the ``context'' of other text information management tasks, including ad hoc text retrieval and web search, we further prove the effectiveness of contextual text mining techniques in a quantitative way with large scale datasets. The framework of contextual text mining not only unifies many explorations of text analysis with context information, but also opens up many new possibilities for future research directions in text mining.

Inclusão e deficiência: em busca das representações sociais na mídia impressa em Natal/RN

This thesis is the result of an extensive research on print media discourse on the inclusion of disabled people in society. Articles published in the newspapers Diário de Natal/O Poti and Tribuna do Norte from 1992 to 2002 have been analyzed. Beginning with the very same questions that moved Moscovici (1978) in his classic study La Psychalyse son image et son public , according to which the media play a predominant role in the formation and propagation of social representations as well as in the construction of human behaviors, we have chosen this mass media as our investigation field. Understanding the importance of the communication theory, we intend to associate it to the social representation theory, since communication, as an aspiration, relates to the fundamentals of all humanity (WOLTON, 2004, p. 56). Moreover, means of communication represent an important space for symbolic production and representational process, allowing the analysis of the circulating discourses on social inclusion and disability. Based on these questions, we have determined social representations present in print media on the subject to be our study object. This objective was elected due to the fact that the thematic of disability and inclusion is scarcely and sporadically found in journalistic speech. The research questions have been: which is the representation of disabled people s condition in print media? What changes have occurred during the analyzed period and which was the role played by print media in this process? The research corpus was composed of newspaper articles about various aspects concerning disability and of free word association by reporters. We have analysed: 1) graphical language promoted by the picture of the substances propagated in the period from 1992 a 2002; 2) free word association experiments carried out with reporters of both newspapers; and 3) texts published from 1996 to 1997 using the high-tech program ALCESTE (Contextual Lexical Analysis of a Set of Segments of Texts). The results revealed that the print media in Natal/RN refer to the topic in a discontinuous way, and depend on specific events to highlight disabled people s fight for their rights. Social inclusion is still a great challenge for these people in all levels. We believe that this incapacity to overcome all kinds of obstacles is established in a dialectic relation between society and the media: society remains silent (the manifestation of interest for the cause only attracts some people s or groups attention) and the media, which selects which information is to be broadcasted, gives no evidence to the issue. This representation may be noticed in the infrequency in which articles about the subject are published, as well as in the emphasis to sports, a more important issue for the media. An implication of this study is that a new perspective is opened for analysis and reflection: the Paralympics games as both an inclusive and a segregating social phenomenon. It would be beautiful to have all of us together!

Semantic sentiment analysis of Twitter

Sentiment analysis over Twitter offer organisations a fast and effective way to monitor the publics' feelings towards their brand, business, directors, etc. A wide range of features and methods for training sentiment classifiers for Twitter datasets have been researched in recent years with varying results. In this paper, we introduce a novel approach of adding semantics as additional features into the training set for sentiment analysis. For each extracted entity (e.g. iPhone) from tweets, we add its semantic concept (e.g. Apple product) as an additional feature, and measure the correlation of the representative concept with negative/positive sentiment. We apply this approach to predict sentiment for three different Twitter datasets. Our results show an average increase of F harmonic accuracy score for identifying both negative and positive sentiment of around 6.5% and 4.8% over the baselines of unigrams and part-of-speech features respectively. We also compare against an approach based on sentiment-bearing topic analysis, and find that semantic features produce better Recall and F score when classifying negative sentiment, and better Precision with lower Recall and F score in positive sentiment classification.

Dynamic joint sentiment-topic model

Social media data are produced continuously by a large and uncontrolled number of users. The dynamic nature of such data requires the sentiment and topic analysis model to be also dynamically updated, capturing the most recent language use of sentiments and topics in text. We propose a dynamic Joint Sentiment-Topic model (dJST) which allows the detection and tracking of views of current and recurrent interests and shifts in topic and sentiment. Both topic and sentiment dynamics are captured by assuming that the current sentiment-topic-specific word distributions are generated according to the word distributions at previous epochs. We study three different ways of accounting for such dependency information: (1) Sliding window where the current sentiment-topic word distributions are dependent on the previous sentiment-topic-specific word distributions in the last S epochs; (2) skip model where history sentiment topic word distributions are considered by skipping some epochs in between; and (3) multiscale model where previous long- and shorttimescale distributions are taken into consideration. We derive efficient online inference procedures to sequentially update the model with newly arrived data and show the effectiveness of our proposed model on the Mozilla add-on reviews crawled between 2007 and 2011. © 2013 ACM 2157-6904/2013/12-ART5 $ 15.00.

Análise dos métodos de medição dos parâmetros geométricos de via e correlação entre os dados obtidos

A razão atribuída à escolha do tema do T.F.M. em Caminhos-de-ferro, tem a ver com o facto de ser uma via de comunicação específica. Contrariamente ao que acontece na execução de auto-estradas, no caminho-de-ferro, a gestora da infra-estrutura responsabiliza-se pela execução, exploração (sinalização e controlo de circulação), abastecimento de energia fornecida às vias que dispõem de catenária (tracção eléctrica), manutenção e conservação da via. O processo de análise e inspecção de geometria de via, é usado quando e necessário preservar a infra-estrutura. Este processo iniciou-se nos caminhos-de-ferro portugueses, há muitos anos, depois da inauguração do primeiro troco de linha férrea. A primeira viagem ocorre em Outubro de 1856, sendo o início do processo em 1968, com a dresina “Matisa PV-6”. Em 1991 a C.P. adquiriu outro veículo de inspecção de via, tendo sido escolhido o VIV02 EM 120 da marca Plasser & Theurer, para substituir “Matisa PV-6”. O tema Análise dos Métodos de Medição dos Parâmetros Geométricos de Via e Correlação entre os Dados Obtidos está directamente relacionado com a manutenção e conservação de via. Na Unidade Operacional Sul (hoje ROS – Região Operacional Sul), local onde desenvolvi o T.F.M., não existem obras de construção de caminhos-de-ferro que pudesse acompanhar e constituir tema para o meu trabalho. Na realidade, com a falta de investimento que se perspectiva no futuro próximo, a manutenção da infra-estrutura passa a ser a actividade principal desenvolvida pela REFER, de modo a assegurar a comodidade, segurança e rapidez na deslocação de cargas e pessoas. A Analise Geométrica de Via e actualmente uma das principais actividades no âmbito da manutenção, que é feita por diagnóstico, contrariamente ao que acontecia no passado em que a conservação metódica era realizada num determinado ano num troço seleccionado independentemente da necessidade ou não da mesma. Uma ajuda preciosa, no que se refere à decisão de se realizar um determinado trabalho de conservação, e a do veículo VIV02 EM 120 que faz inspeções ao longo de toda a rede ferroviária e permite recolher dados e classificar através do desvio padrão, troços com extensão de 200 metros, obtendo os dados relevantes sobre a necessidade de intervenção. Para além do referido veículo existem também equipamentos ligeiros de inspecção dos parâmetros geométricos de via. Um desses equipamentos designa-se por Trólei, não sendo motorizado, pois o mesmo é movido manualmente por um operador. Obviamente que este equipamento não faz a inspecção a toda a rede pois a operação de medição é morosa, sendo contudo utilizado para análise de defeitos geométricos em pequenos trocos, tornando-se assim uma mais-valia, evitando o deslocar de um equipamento “pesado” como o VIV 02 EM 120. Para atingir os objectivos deste trabalho realizaram-se testes de medição com ambos (veiculo e equipamento ligeiro), no mesmo espaço temporal e com as mesmas características físicas, como a temperatura, humidade etc. Os resultados, de acordo com os objectivos, são a comparação entre as medições de ambos, com vista a comprovar a sua utilidade e necessidade, de acordo com os vários tipos de superstruturas constituintes da rede ferroviária nacional.

Discurso mediático e sociedade: repensar a Análise Crítica do Discurso

A análise do discurso jornalístico e do seu enraizamento social tem conhecido avanços significativos nas últimas duas décadas, especialmente devido ao surgimento e desenvolvimento da Análise Crítica do Discurso. No entanto, há três aspectos importantes que merecem mais investigação: o plano temporal na análise do discurso, as estratégias discursivas dos atores sociais, e os efeitos extra e supra-textual do discurso mediatizado. Em primeiro lugar, a compreensão da biografia dos assuntos públicos exige uma análise longitudinal dos textos mediatizados e dos seus contextos sociais, mas a maioria das formas de análise do discurso jornalístico não tem em conta a sequência temporal dos textos e as suas implicações. Em segundo lugar, como a representação mediática das questões sociais é, em grande medida, função da construção discursiva de eventos, problemas e posições por diferentes atores sociais, as estratégias discursivas que eles empregam numa variedade de arenas e canais ‘antes’ e ‘depois’ dos textos jornalísticos precisam de ser examinados. Em terceiro lugar, o facto de que muitos dos modos de operação do discurso são extra- ou supra-textuais requer que se tenha em consideração vários processos sociais ‘fora’ do texto. Este trabalho tem como objetivo produzir um contributo teórico e metodológico para a integração destas questões em análise do discurso, propondo um quadro analítico que combina uma dimensão textual com uma contextual.

Prognostic factors in left-sided endocarditis: results from the andalusian multicenter cohort

Despite medical advances, mortality in infective endocarditis (IE) is still very high. Previous studies on prognosis in IE have observed conflicting results. The aim of this study was to identify predictors of in-hospital mortality in a large multicenter cohort of left-sided IE.Methods An observational multicenter study was conducted from January 1984 to December 2006 in seven hospitals in Andalusia, Spain. Seven hundred and five left-side IE patients were included. The main outcome measure was in-hospital mortality. Several prognostic factors were analysed by univariate tests and then by multilogistic regression model. Results.The overall mortality was 29.5% (25.5% from 1984 to 1995 and 31.9% from 1996 to 2006; Odds Ratio 1.25; 95% Confidence Interval: 0.97-1.60; p = 0.07). In univariate analysis, age, comorbidity, especially chronic liver disease, prosthetic valve, virulent microorganism such as Staphylococcus aureus, Streptococcus agalactiae and fungi, and complications (septic shock, severe heart failure, renal insufficiency, neurologic manifestations and perivalvular extension) were related with higher mortality. Independent factors for mortality in multivariate analysis were: Charlson comorbidity score (OR: 1.2; 95% CI: 1.1-1.3), prosthetic endocarditis (OR: 1.9; CI: 1.2-3.1), Staphylococcus aureus aetiology (OR: 2.1; CI: 1.3-3.5), severe heart failure (OR: 5.4; CI: 3.3-8.8), neurologic manifestations (OR: 1.9; CI: 1.2-2.9), septic shock (OR: 4.2; CI: 2.3-7.7), perivalvular extension (OR: 2.4; CI: 1.3-4.5) and acute renal failure (OR: 1.69; CI: 1.0-2.6). Conversely, Streptococcus viridans group etiology (OR: 0.4; CI: 0.2-0.7) and surgical treatment (OR: 0.5; CI: 0.3-0.8) were protective factors.Conclusions Several characteristics of left-sided endocarditis enable selection of a patient group at higher risk of mortality. This group may benefit from more specialised attention in referral centers and should help to identify those patients who might benefit from more aggressive diagnostic and/or therapeutic procedures.

Three-year multicenter surveillance of community-acquired Listeria monocytogenes meningitis in adults

BACKGROUND. Listeria monocytogenes is the third most frequent cause of bacterial meningitis. The aim of this study is to know the incidence and risk factors associated with development of acute community-acquired Lm meningitis in adult patients and to evaluate the clinical features, management, and outcome in this prospective case series. METHODS. A descriptive, prospective, and multicentric study carried out in 9 hospitals in the Spanish Network for Research in Infectious Diseases (REIPI) over a 39-month period. All adults patients admitted to the participating hospitals with the diagnosis of acute community-acquired bacterial meningitis (Ac-ABM) were included in this study. All these cases were diagnosed on the basis of a compatible clinical picture and a positive cerebrospinal fluid (CSF) culture or blood culture. The patients were followed up until death or discharge from hospital. RESULTS. Two hundred and seventy-eight patients with Ac-ABM were included. Forty-six episodes of Lm meningitis were identified in 46 adult patients. In the multivariate analysis only age (OR 1.026; 95% CI 1.00-1.05; p = 0.042), immunosuppression (OR 2.520; 95% CI 1.05-6.00; p = 0.037), and CSF/blood glucose ratio (OR 39.42; 95% CI 4.01-387.50; p = 0.002) were independently associated with a Lm meningitis. The classic triad of fever, neck stiffness and altered mental status was present in 21 (49%) patients, 32% had focal neurological findings at presentation, 12% presented cerebellum dysfunction, and 9% had seizures. Twenty-nine (68%) patients were immunocompromised. Empirical antimicrobial therapy was intravenous ampicillin for 34 (79%) of 43 patients, in 11 (32%) of them associated to aminoglycosides. Definitive ampicillin plus gentamicin therapy was significantly associated with unfavourable outcome (67% vs 28%; p = 0.024) and a higher mortality (67% vs 32%; p = 0.040).The mortality rate was 28% (12 of 43 patients) and 5 of 31 (16.1%) surviving patients developed adverse clinical outcome. CONCLUSIONS Elderly or immunocompromised patients, and a higher CSF/blood glucose ratio in patients with Ac-ABM must alert clinicians about Lm aetiology. Furthermore, we observed a high incidence of acute community-acquired Lm meningitis in adults and the addition of aminoglycosides to treatment should be avoid in order to improve the patients' outcome. Nevertheless, despite developments in intensive care and antimicrobial therapy, this entity is still a serious disease that carries high morbidity and mortality rates.

Ki-67 MIB1 labelling index and the prognosis of primary TaT1 urothelial cell carcinoma of the bladder.

Aims: To evaluate whether ki-67 labelling index (LI) has independent prognostic value for survival of patients with bladder urothelial tumours graded according to the 2004 World Health Organisation classification. Methods: Ki-67 LI was evaluated in 164 cases using the grid counting method. Non-invasive (stage Ta) tumours were: papilloma (n = 5), papillary urothelial neoplasia of low malignant potential (PUNLMP; n = 26), and low (LG; n = 34) or high grade (HG; n = 15) papillary urothelial carcinoma. Early invasive (stage T1) tumours were: LG (n = 58) and HG (n = 26) carcinoma. Statistical analysis included Fisher and x2 tests, and mean comparisons by ANOVA and t test. Univariate and multivariate survival analyses were performed according to the Kaplan–Meier method with log rank test and Cox’s proportional hazard method. Results: Mean ki-67 LI increased from papilloma to PUNLMP, LG, and HG in stage Ta (p,0.0001) and from LG to HG in stage T1 (p = 0.013) tumours. High tumour proliferation (.13%) was related to greater tumour size (p = 0.036), recurrence (p = 0.036), progression (p = 0.035), survival (p = 0.054), and high p53 accumulation (p = 0.015). Ki-67 LI and tumour size were independent predictors of disease free survival (DFS), but only ki-67 LI was related to progression free survival (PFS). Cancer specific overall survival (OS) was related to ki-67 LI, tumour size, and p27kip1 downregulation. Ki-67 LI was the main independent predictor of DFS (p = 0.0005), PFS (p = 0.0162), and cancer specific OS (p = 00195). Conclusion: Tumour proliferation measured by Ki-67 LI is related to tumour recurrence, stage progression, and is an independent predictor of DFS, PFS, and cancer specific OS in TaT1 bladder urothelial cell carcinoma.

Deep-brain stimulation for Parkinson's disease

Comment on N Engl J Med. 2010 Jun 3;362(22):2077-91 author reply 988.

Intratracheal dopamine attenuates pulmonary edema and improves survival after ventilator-induced lung injury in rats

INTRODUCTION Clearance of alveolar oedema depends on active transport of sodium across the alveolar-epithelial barrier. beta-Adrenergic agonists increase clearance of pulmonary oedema, but it has not been established whether beta-agonist stimulation achieves sufficient oedema clearance to improve survival in animals. The objective of this study was to determine whether the increased pulmonary oedema clearance produced by intratracheal dopamine improves the survival of rats after mechanical ventilation with high tidal volume (HVT). METHODS This was a randomized, controlled, experimental study. One hundred and thirty-two Wistar-Kyoto rats, weighing 250 to 300 g, were anaesthetized and cannulated via endotracheal tube. Pulmonary oedema was induced by endotracheal instillation of saline solution and mechanical ventilation with HVT. Two types of experiment were carried out. The first was an analysis of pulmonary oedema conducted in six groups of 10 rats ventilated with low (8 ml/kg) or high (25 ml/kg) tidal volume for 30 or 60 minutes with or without intratracheally instilled dopamine. At the end of the experiment the animals were exsanguinated and pulmonary oedema analysis performed. The second experiment was a survival analysis, which was conducted in two groups of 36 animals ventilated with HVT for 60 minutes with or without intratracheal dopamine; survival of the animals was monitored for up to 7 days after extubation. RESULTS In animals ventilated at HVT with or without intratracheal dopamine, oxygen saturation deteriorated over time and was significantly higher at 30 minutes than at 60 minutes. After 60 minutes, a lower wet weight/dry weight ratio was observed in rats ventilated with HVT and instilled with dopamine than in rats ventilated with HVT without dopamine (3.9 +/- 0.27 versus 4.9 +/- 0.29; P = 0.014). Survival was significantly (P = 0.013) higher in animals receiving intratracheal dopamine and ventilated with HVT, especially at 15 minutes after extubation, when 11 of the 36 animals in the HVT group had died as compared with only one out of the 36 animals in the HVT plus dopamine group. CONCLUSION Intratracheal dopamine instillation increased pulmonary oedema clearance in rats ventilated with HVT, and this greater clearance was associated with improved survival.

Dynamic arterial elastance to predict arterial pressure response to volume loading in preload-dependent patients

INTRODUCTION Hemodynamic resuscitation should be aimed at achieving not only adequate cardiac output but also sufficient mean arterial pressure (MAP) to guarantee adequate tissue perfusion pressure. Since the arterial pressure response to volume expansion (VE) depends on arterial tone, knowing whether a patient is preload-dependent provides only a partial solution to the problem. The objective of this study was to assess the ability of a functional evaluation of arterial tone by dynamic arterial elastance (Ea(dyn)), defined as the pulse pressure variation (PPV) to stroke volume variation (SVV) ratio, to predict the hemodynamic response in MAP to fluid administration in hypotensive, preload-dependent patients with acute circulatory failure. METHODS We performed a prospective clinical study in an adult medical/surgical intensive care unit in a tertiary care teaching hospital, including 25 patients with controlled mechanical ventilation who were monitored with the Vigileo(®) monitor, for whom the decision to give fluids was made because of the presence of acute circulatory failure, including arterial hypotension (MAP ≤65 mmHg or systolic arterial pressure <90 mmHg) and preserved preload responsiveness condition, defined as a SVV value ≥10%. RESULTS Before fluid infusion, Ea(dyn) was significantly different between MAP responders (MAP increase ≥15% after VE) and MAP nonresponders. VE-induced increases in MAP were strongly correlated with baseline Ea(dyn) (r(2) = 0.83; P < 0.0001). The only predictor of MAP increase was Ea(dyn) (area under the curve, 0.986 ± 0.02; 95% confidence interval (CI), 0.84-1). A baseline Ea(dyn) value >0.89 predicted a MAP increase after fluid administration with a sensitivity of 93.75% (95% CI, 69.8%-99.8%) and a specificity of 100% (95% CI, 66.4%-100%). CONCLUSIONS Functional assessment of arterial tone by Ea(dyn), measured as the PVV to SVV ratio, predicted arterial pressure response after volume loading in hypotensive, preload-dependent patients under controlled mechanical ventilation.

Confabulation in schizophrenia: a neuropsychological study.

Confabulation has been documented in schizophrenia, but its neuropsychological correlates appear to be different from those of confabulation in neurological disease states. Forty-five schizophrenic patients and 37 controls were administered a task requiring them to recall fables. They also underwent testing with a range of memory and executive tasks. The patients with schizophrenia produced significantly more confabulations than the controls. After correcting for multiple comparisons, confabulation was not significantly associated with memory impairment, and was associated with impairment on only one of eight executive measures, the Brixton Test. Confabulation scores were also associated with impairment on two semantic memory tests. Confabulation was correlated with intrusion errors in recall, but not false positive errors in a recognition task. The findings suggest that confabulation in schizophrenia is unrelated to the episodic memory impairment seen in the disorder. However, the association with a circumscribed deficit in executive function could be consistent with a defective strategic retrieval account of confabulation similar to that of Moscovitch and co-workers, interacting with defective semantic memory.

Angiogenesis-related gene expression profile with independent prognostic value in advanced ovarian carcinoma.

BACKGROUND Ovarian carcinoma is the most important cause of gynecological cancer-related mortality in Western societies. Despite the improved median overall survival in patients receiving chemotherapy regimens such as paclitaxel and carboplatin combination, relapse still occurs in most advanced diseased patients. Increased angiogenesis is associated with rapid recurrence and decreased survival in ovarian cancer. This study was planned to identify an angiogenesis-related gene expression profile with prognostic value in advanced ovarian carcinoma patients. METHODOLOGY/PRINCIPAL FINDINGS RNAs were collected from formalin-fixed paraffin-embedded samples of 61 patients with III/IV FIGO stage ovarian cancer who underwent surgical cytoreduction and received a carboplatin plus paclitaxel regimen. Expression levels of 82 angiogenesis related genes were measured by quantitative real-time polymerase chain reaction using TaqMan low-density arrays. A 34-gene-profile which was able to predict the overall survival of ovarian carcinoma patients was identified. After a leave-one-out cross validation, the profile distinguished two groups of patients with different outcomes. Median overall survival and progression-free survival for the high risk group was 28.3 and 15.0 months, respectively, and was not reached by patients in the low risk group at the end of follow-up. Moreover, the profile maintained an independent prognostic value in the multivariate analysis. The hazard ratio for death was 2.3 (95% CI, 1.5 to 3.2; p<0.001). CONCLUSIONS/SIGNIFICANCE It is possible to generate a prognostic model for advanced ovarian carcinoma based on angiogenesis-related genes using formalin-fixed paraffin-embedded samples. The present results are consistent with the increasing weight of angiogenesis genes in the prognosis of ovarian carcinoma.