No description available.

The Digital Library Research Laboratory (DLRL) has collected over 3.5 billion tweets on different events for the Coordinated, Behaviorally-Aware Recovery for Transportation and Power Disruptions (CBAR-tpd), the Integrated Digital Event Archiving and Library (IDEAL), and the Global Event Trend Archive Research (GETAR) projects. The tweet collection topics include heart attack, solar eclipse, terrorism, etc. There are several collections on naturally occurring events such as hurricanes, floods, and solar eclipses. Such naturally occurring events are distributed across space and time. It would be beneficial to researchers if we can perform a spatial-temporal analysis to test some hypotheses, and to find any trends that tweets would reveal for such events. I apply an existing algorithm to detect locations from tweets by modifying it to work better with the type of datasets I work with. I use the time captured in tweets and also identify the tense of the sentences in tweets to perform the temporal analysis. I build a rule-based model for obtaining the tense of a tweet. The results from these two algorithms are merged to analyze naturally occurring moving events such as solar eclipses and hurricanes. Using the spatial-temporal information from tweets, I study if tweets can be a relevant source of information in understanding the movement of the event. I create visualizations to compare the actual path of the event with the information extracted by my algorithms. After examining the results from the analysis, I noted that Twitter can be a reliable source to identify places affected by moving events almost immediately. The locations obtained are at a more detailed level than in news-wires. We can also identify the time that an event affected a particular region by date. / Master of Science / News now travels faster on social media than through news channels. Information from social media can help retrieve minute details that might not be emphasized in news. People tend to describe their actions or sentiments in tweets. I aim at studying if such collections of tweets are dependable sources for identifying paths of moving events. In events like hurricanes, using Twitter can help in analyzing people’s reaction to such moving events. These may include actions such as dislocation or emotions during different phases of the event. The results obtained in the experiments concur with the actual path of the events with respect to the regions affected and time. The frequency of tweets increases during event peaks. The number of locations affected that are identified are significantly more than in news wires.

Last decade witnessed the rise of online social networks, especially Twitter. Today, Twitteris a giant social platform with over 250 million users |who produce massive amounts of data everyday. This creates many research opportunities, speci cally for Natural Language Processing (NLP) in which text is utilized to extract information that could be used in many applications. One problem NLP might help solving is geolocation inference or geolocation detection from online social networks. Detecting the location of Twitter users based on the text of their tweets is useful since not many users publicly declare their locations or geotag their tweets. Location information is crucial for a variety of applications such as event detection, disease and illness tracking and user pro ling. These tasks are not trivial, because online content is often noisy; it includes misspellings, incomplete words or phrases, idiomatic expressions, abbreviations, acronyms, and Twitter-speci c literature. In this work, we attempted to detect the location of Canadian users |and tweets sent from Canada |at metropolitan areas and province level; this was not done before, to the best of our knowledge. In order to do this, we collected two di erent datasets, and applied a variety of machine learning, including deep learning methods. Besides, we also attempted to geolocate users based on their social graph (i.e., user's friends and followers) as a novel approach.

Submitted by Tatiana Lima (tatianasl@ufba.br) on 2016-08-30T21:05:13Z No. of bitstreams: 1 Amaral, Marcelo Santos.pdf: 5121263 bytes, checksum: 352e9409f1f2a35901488dc5bbf1474d (MD5) / Approved for entry into archive by Maria Angela Dortas (dortas@ufba.br) on 2016-08-30T21:13:36Z (GMT) No. of bitstreams: 1 Amaral, Marcelo Santos.pdf: 5121263 bytes, checksum: 352e9409f1f2a35901488dc5bbf1474d (MD5) / Made available in DSpace on 2016-08-30T21:13:36Z (GMT). No. of bitstreams: 1 Amaral, Marcelo Santos.pdf: 5121263 bytes, checksum: 352e9409f1f2a35901488dc5bbf1474d (MD5) / O surgimento de novas tecnologias da informação e da comunicação motivou estudos sobre as possíveis transformações provocadas por estas tecnologias, fazendo surgir diferentes correntes e dividindo os teóricos em moderados, otimistas e pessimistas em relação a estas mudanças. Mais recentemente, o advento das novas mídias sociais da internet também têm motivado diferentes investigações sobre os possíveis impactos que elas possam ter sobre as relações sociais, econômicas e políticas. Instituições, organizações, políticos e cidadãos estão adotando o uso das redes sociais em ritmo acelerado, levantando questões sobre os efeitos que esta adoção pode provocar nas relações entre Estado, governo e sociedade. O Twitter é um serviço de microblogging que surgiu com o propósito de ser um diário público de atividades, mas foi apropriado e utilizado de forma bem mais ampla, tornando-se uma rede social única na internet. Este estudo exploratório investigou a adoção do Twitter pelo Congresso Nacional e sua utilização durante o período eleitoral de 2014, abrangendo quatro momentos de análises: um período prévio ao momento eleitoral, o início e o final da campanha política, e um período posterior às eleições de 2014. Os indicadores analisados incluem dados sócio-políticos dos parlamentares em exercício em qualquer das quatro datas de análise e seus dados de uso do Twitter, coletados diretamente da rede social. O objetivo foi descrever e analisar como parlamentares e partidos políticos representados no Congresso Nacional atuavam no Twitter, antes, durante e após a campanha eleitoral de 2014. Os dados indicam que os parlamentares aderiram ao Twitter antes das eleições de 2010 ou antes das eleições de 2014. Em momentos de eleição, tendem a aumentar suas presenças na rede, postando mais conteúdos e buscando visibilidade para suas campanhas eleitorais, obtendo com isto mais popularidade e, dentre os mais populares, mais votos. Há parlamentares que, por sua grande popularidade no Twitter, se equivalem a celebridades virtuais, sendo excepcionalmente seguidos por colegas e por cidadãos, mas seguindo poucas pessoas em retorno. Na outra ponta, há uma parcela de parlamentares que se mostraram indiferentes ao uso do recurso durante as quatro etapas de análise. Os partidos políticos também reproduzem as relações do ambiente off-line no Twitter, tendo os principais partidos liderado o movimento de adesão. Há partidos, no entanto, que possuem importância relativamente maior no ambiente virtual do que fora dele, sendo muito seguidos por políticos de outros partidos e pelo público externo ao Congresso. Os parlamentares aumentaram ignificativamente o número de seus seguidores entre o início e o final do período analisado, mas o público dos partidos políticos não ficou mais qualificado com este acréscimo de popularidade de seus representantes. Depois das eleições, muitos parlamentares abandonam o uso da ferramenta – especialmente aqueles que não obtiveram sucesso eleitoral; outros, reeleitos, diminuíram suas presenças no ambiente virtual para um ritmo próximo do anterior. As conclusões apontam para a reprodução das relações personalistas e individualistas da política brasileira no ambiente virtual, com as conexões entre cidadãos e políticos valorizando a figura dos políticos, em detrimento das legendas políticas às quais estão filiados. / The rise of new information technologies and communication has motivated research on the potential changes caused by these technologies, giving rise to different streams and splitting the theoretical in moderate, optimistic and pessimistic about these changes. Recently, the advent of new social media of the Internet also have motivated different investigations on the potential impacts they might have on the social, economic and political relations. Institutions, organizations, politicians and citizens are adopting the use of social networks at a fast pace, raising questions about the effects that adoption may result in the relations between state, government and society. Twitter is a microblogging service that has emerged intended to be a public microblogging, but was appropriated and used far more widely becoming a unique social network on the Internet. This exploratory study has investigated the adoption of Twitter by the National Congress and its use during the Brazilian election of 2014, covering four stages of analysis: one period prior to election time, the beginning and the end of the political campaign, and a period after the 2014 elections. The indicators examined here include sociopolitical data of parliamentarians in any of the four dates of analysis and their Twitter usage data collected directly from the social network. The aimed to describe and analyze how parliamentarians and political parties represented in the National Congress acted on Twitter before, during and after the election campaign of 2014. The data indicate that parliamentarians have joined Twitter before the 2010 elections or before the 2014 elections. In moments of elections, they tend to increase their presence on the network, posting more content and seeking visibility for their election campaigns, obtaining with this more popularity and among the most popular, more votes. There are parliamentarians, for their great popularity on Twitter, are equivalent to virtual celebrities, being exceptionally followed by colleagues and citizens, but following few people in return. At the other hand, a part of parliamentarians have shown indifferent to the use of the resource during the four stages of analysis. Political parties also reproduce relations of the offline environment on Twitter, with major parties led the membership movement. There are parties, however, which have relatively more important in the virtual environment than outside it, being very followed by politicians from other parties and the external public to Congress. Parliamentarians have significantly increased the number of its followers between the beginning and the end of the reporting period, but the audience of political parties was not more qualified with this popularity increase their representatives. After the elections, many parliaments have abandoned the use of the tool – especially those who did not achieve electoral success; others, re-elected, decreased their presence in the virtual environment. The findings point to the reproduction of personalist and individualist relations in Brazilian politics in the virtual environment, with the connections between citizens and politicians valuing the figure of politicians at the expense of political parties to which they are affiliated.

Cette thèse présente une étude sur la mesure des réseaux sociaux en ligne avec un accent particulier sur Twitter qui est l'un des plus grands réseaux sociaux. Twitter utilise exclusivement des liens dirigés entre les comptes. Cela rend le graphe social de Twitter beaucoup plus proche que Facebok du graphe social représentant les communications dans la vie réelle. Par conséquent, la compréhension de la structure du graphe social de Twitter et de la manière dont les informations se propagent dans le graphe est intéressant non seulement pour les informaticiens, mais aussi pour les chercheurs dans d'autres domaines, tels que la sociologie. Cependant, on sait peu de choses sur la propagation de l'information sur Twitter. / This thesis presents the measurement study of Online Social Networks focusing on Twitter. Twitter is one of the largest social networks using exclusively directed links among accounts. This makes the Twitter social graph much closer to the social graph supporting real life communications than, for instance, Facebook. Therefore, understanding the structure of the Twitter social graph and the way information propagates through it is interesting not only for computer scientists, but also for researchers in other fields, such as sociologists. However, littles is known about the information propagation in Twitter. In the first part, we present an in-depth study of the macroscopic structure of the Twitter social graph. In the second part, we study the propagation of the news media articles shared on Twitter. In the third part we present an experimental study of graph sampling.

Magíster en Ciencias, Mención Computación / Motivación: El problema de manejar grandes cantidades de datos producidos por usuarios de los llamados sitios de social media ya no parece ser nuevo. Por ejemplo, en la red social online Twitter cada día se publica más de 400 millones de mensajes. Y este diluvio de datos está afectando a cada vez más escenarios. En particular, el problema de comprender qué está pasando en el mundo se vuelve un problema cada vez más difícil, debido a la gran cantidad de fuentes de noticias. Breaking News corresponde a noticias que interrumpen el flujo normal de radio y televisión. En las redes sociales son un fenómeno más complicado de estudiar, debido a un paradigma distinto: la producción y el consumo descentralizado de datos en modo streaming. Comprender el impacto de las noticias en las redes sociales online es un problema difícil e interesante de investigar. Propósito: El objetivo principal de este proyecto es responder a la pregunta: ¿Es posible predecir si un evento noticioso causará impacto en las redes sociales online, poco después de su publicación? Se utiliza Twitter como la fuente de datos. Contribuciones: Se define la noción de impacto, basado en la tasa de llegada de los tweets que discuten los eventos. Luego, se predice el impacto de los eventos usando características de los mensajes, como el número de veces que son compartidos, el puntaje de sentimiento, etc. La tarea de predicción tiene buenos resultados de clasificación incluso usando el primer 5% de los datos, y aunque mejora al agregar más datos, F-score y accuracy decrecen al usar el 100%. Además, se caracterizan los eventos en Twitter, encontrando propiedades distintivas entre eventos de alto y bajo impacto. En los eventos de alto impacto la información se propaga a una mayor velocidad y escala. Además, son más focalizados en términos de vocabulario, y muestran mayor puntaje de sentimiento. Eventos de bajo impacto son más conversacionales: los usuarios que hablan acerca de esos eventos usan un vocabulario más extenso y comparten más recursos, como URLs o hashtags. Metodología: Se propone y desarrolla una metodología de recolección de datos e identificación de eventos. El conjunto de datos final contiene 9,000 eventos y 45 millones de tweets, aproximadamente. Para asignar la categoría de impacto de un evento, se usan métodos de clustering para aprender una representación multidimensional de éstos. Luego, se distingue entre eventos de alto y bajo impacto. Usando un clasificador de regresión logística sobre porciones de los datos, se clasifican eventos para predecir su categoría de impacto, usando la representación multidimensional como base para la evaluación. Valor: El valor de este trabajo yace en sus posibles aplicaciones: puede apoyar el trabajo periodístico, sirve para generar resúmenes automáticos valiosos, desarrollar sistemas de recomendación, publicidad focalizada, encontrar contenido relevante, entre otras.

Seminario para optar al grado de Ingeniero Comercial, Mención Administración / La forma de entender y usar los medios ha cambiado significativamente en las últimas décadas. Gracias al vertiginoso desarrollo de las tecnologías de la información han nacido y se han consolidado medios (Internet) que permiten una interacción bidireccional los cuales se han acoplado más que reemplazado al uso de medios tradicionales en diversas formas, una de las cuales se da a través de los comentarios que realizan los espectadores a través Twitter o Facebook. Por otro lado, existe en la actualidad el debate sobre la eliminación o no del People Meter © Online, herramienta que permite medir las audiencias en tiempo real, lo que puede cambiar el cómo funciona la industria tal y como la conocemos. En base a esta inquietud el presente estudio examinó como es la relación entre el nivel de audiencia y la cantidad de comentarios acerca de un mismo programa en la red social Twitter de manera de determinar si esta medida es una alternativa viable para el Rating Online o se trata de una medición complementaria. Se utilizó el caso del programa chileno Mundos Opuestos de Canal 13, que tuvo un alto nivel de audiencia y se transmitió de manera franjeada en horario prime durante el primer semestre del año 2012, encontrándose importante evidencia. En primer lugar se encontró una correlación significativa y positiva alta entre el nivel de audiencia y la cantidad de tweets sobre el programa a nivel total y en casi todos los segmentos, entregando evidencia empírica sobre Twitter como un espejo de lo que es contingente a nivel de sociedad. En segundo lugar se encontró que existe una relación temporal a nivel diario significativa para casi todos los segmentos entre la cantidad de tweets de la emisión anterior y el rating actual, y viceversa. Sin embargo la relación tiene un grado de ajuste bajo. Dado lo anterior se puede afirmar que la cantidad de tweets puede ser considerada una métrica paralela que está asociada al rating online al menos en un programa de alta popularidad. Sin embargo, no se puede afirmar que ambas midan exactamente el mismo fenómeno. Es posible hipotetizar que, más que preferencia sobre un programa, evalúa su popularidad en términos de las apreciaciones de los distintos espectadores a partir de lo que está ocurriendo en el programa. Dicho de otro modo, el uso de Twitter puede ser usado más bien como una medida de impacto social del programa televisivo.

Tese de Mestrado Integrado. Engenharia Informática e Computação. Faculdade de Engenharia. Universidade do Porto. 2012

Tese de mestrado integrado. Engenharia Informática e Computação. Faculdade de Engenharia. Universidade do Porto. 2010

In 2011, Internet users spent almost 23% of their time on social media sites such as Twitter and Facebook. Twitter alone was estimated to have over 200 million active users. With social media being such a popular online pastime, a tremendous amount of information becomes available from the posts that users put on social media sites. This information has the potential to reveal details about the social media users, such as the relationship between characteristics of the users and what they post. This relationship is a hot research topic and one of the most frequently studied characteristic is the gender of a user. Feature frequency is often included in such a task, but this thesis shows that for Twitter tweets it either does not contribute significantly to gender classification or hinders classification. / McAnulty College and Graduate School of Liberal Arts; / Computational Mathematics / MS; / Thesis;