The ongoing European Refugee Crisis has been one of the most popular trending topics on Twitter for the past 8 months. This paper applies topic modeling on bulks of tweets to discover the hidden patterns within these social media discussions. In particular, we perform topic analysis through solving Non-negative Matrix Factorization (NMF) as an Inexact Alternating Least Squares problem. We accelerate the computation using techniques including tweet sampling and augmented NMF, compare NMF results with different ranks and visualize the outputs through topic representation and frequency plots. We observe that supportive sentiments maintained a strong presence while negative sentiments such as safety concerns have emerged over time.

Thesis (PhD (Industrial Engineering))--University of Stellenbosch, 2010. / ENGLISH ABSTRACT: The crucial role of innovation in creating sustainable competitive advantage is widely recognised in industry today. Likewise, the importance of having the required information accessible to the right employees at the right time is well-appreciated. More specifically, the dependency of effective, efficient innovation processes on the availability of information has been pointed out in literature. A great challenge is countering the effects of the information overload phenomenon in organisations in order for employees to find the information appropriate to their needs without having to wade through excessively large quantities of information to do so. The initial stages of the innovation process, which are characterised by free association, semi-formal activities, conceptualisation, and experimentation, have already been identified as a key focus area for improving the effectiveness of the entire innovation process. The dependency on information during these early stages of the innovation process is especially high. Any organisation requires a strategy for innovation, a number of well-defined, implemented processes and measures to be able to innovate in an effective and efficient manner and to drive its innovation endeavours. In addition, the organisation requires certain enablers to support its innovation efforts which include certain core competencies, technologies and knowledge. Most importantly for this research, enablers are required to more effectively manage and utilise innovation-related information. Information residing inside and outside the boundaries of the organisation is required to feed the innovation process. The specific sources of such information are numerous. Such information may further be structured or unstructured in nature. However, an ever-increasing ratio of available innovation-related information is of the unstructured type. Examples include the textual content of reports, books, e-mail messages and web pages. This research explores the innovation landscape and typical sources of innovation-related information. In addition, it explores the landscape of text analytical approaches and techniques in search of ways to more effectively and efficiently deal with unstructured, textual information. A framework that can be used to provide a unified, dynamic view of an organisation‟s innovation-related information, both structured and unstructured, is presented. Once implemented, this framework will constitute an innovation-focused knowledge base that will organise and make accessible such innovation-related information to the stakeholders of the innovation process. Two novel, complementary text analytical techniques, Latent Dirichlet Allocation and the Concept-Topic Model, were identified for application with the framework. The potential value of these techniques as part of the information systems that would embody the framework is illustrated. The resulting knowledge base would cause a quantum leap in the accessibility of information and may significantly improve the way innovation is done and managed in the target organisation. / AFRIKAANSE OPSOMMING: Die belangrikheid van innovasie vir die daarstel van „n volhoubare mededingende voordeel word tans wyd erken in baie sektore van die bedryf. Ook die belangrikheid van die toeganklikmaking van relevante inligting aan werknemers op die geskikte tyd, word vandag terdeë besef. Die afhanklikheid van effektiewe, doeltreffende innovasieprosesse op die beskikbaarheid van inligting word deurlopend beklemtoon in die navorsingsliteratuur. „n Groot uitdaging tans is om die oorsake en impak van die inligtingsoorvloedverskynsel in ondernemings te bestry ten einde werknemers in staat te stel om inligting te vind wat voldoen aan hul behoeftes sonder om in die proses deur oormatige groot hoeveelhede inligting te sif. Die aanvanklike stappe van die innovasieproses, gekenmerk deur vrye assosiasie, semi-formele aktiwiteite, konseptualisering en eksperimentasie, is reeds geïdentifiseer as sleutelareas vir die verbetering van die effektiwiteit van die innovasieproses in sy geheel. Die afhanklikheid van hierdie deel van die innovasieproses op inligting is besonder hoog. Om op „n doeltreffende en optimale wyse te innoveer, benodig elke onderneming „n strategie vir innovasie sowel as „n aantal goed gedefinieerde, ontplooide prosesse en metingskriteria om die innovasieaktiwiteite van die onderneming te dryf. Bykomend benodig ondernemings sekere innovasie-ondersteuningsmeganismes wat bepaalde sleutelaanlegde, -tegnologiëe en kennis insluit. Kern tot hierdie navorsing, benodig organisasies ook ondersteuningsmeganismes om hul in staat te stel om meer doeltreffend innovasie-verwante inligting te bestuur en te gebruik. Inligting, gehuisves beide binne en buite die grense van die onderneming, word benodig om die innovasieproses te voer. Die bronne van sulke inligting is veeltallig en hierdie inligting mag gestruktureerd of ongestruktureerd van aard wees. „n Toenemende persentasie van innovasieverwante inligting is egter van die ongestruktureerde tipe, byvoorbeeld die inligting vervat in die tekstuele inhoud van verslae, boeke, e-posboodskappe en webbladsye. In hierdie navorsing word die innovasielandskap asook tipiese bronne van innovasie-verwante inligting verken. Verder word die landskap van teksanalitiese benaderings en -tegnieke ondersoek ten einde maniere te vind om meer doeltreffend en optimaal met ongestruktureerde, tekstuele inligting om te gaan. „n Raamwerk wat aangewend kan word om „n verenigde, dinamiese voorstelling van „n onderneming se innovasieverwante inligting, beide gestruktureerd en ongestruktureerd, te skep word voorgestel. Na afloop van implementasie sal hierdie raamwerk die innovasieverwante inligting van die onderneming organiseer en meer toeganklik maak vir die deelnemers van die innovasieproses. Daar word verslag gelewer oor die aanwending van twee nuwerwetse, komplementêre teksanalitiese tegnieke tot aanvulling van die raamwerk. Voorts word die potensiele waarde van hierdie tegnieke as deel van die inligtingstelsels wat die raamwerk realiseer, verder uitgewys en geillustreer.

Thesis (DLitt (Ancient Studies)) -- University of Stellenbosch, 2004. / The purpose of this study is to investigate the information structure of Biblical Hebrew narrative, and develop accounts of topic and focus in Biblical Hebrew, respectively. Both topic and focus categories have been determined for Biblical Hebrew (cf. chapters 3 and 5), as well as the information structure strategies that these categories can fulfill in discourse. For topic categories, four different categories of information structure topics in Biblical Hebrew have been distinguished. These are 1. Primary topics 2. Secondary topics 3. Tail topics 4. Topic frames In addition, associated with topics are topic contrastiveness as well as deictic orientations or text-world frames. All these categories, when present, are part of the topical framework of a discourse. For focus structure categories, three different types of focus structure in Biblical Hebrew have been distinguished. These are: 1. Predicate focus 2. Sentence focus 3. Argument focus Again, like in the case of topics, contrastiveness is associated with focus structures. The strategies of information structure topics and focus structures in theme developments were distinguished. For topics, the following information structure strategies or functions stand out: 1. Topic continuity 2. Topic promotion 3. Topic shift 4. Topic deictic text-world framing 5. Topic contrasting For focus structures, the following information structure strategies or functions stand out: 1. Commenting on topics 2. Presenting unidentifiable or inactive participants 3. Reporting, that is, event-reporting and state-reporting of out-of-the-blue, unexpected, discourse new events or states. Some reporting re-directs the theme, other reporting, especially that of states, supports the theme. 4. Identifying referents, either as identifying contrastive, unexpected referents or deictic text-world frames, or by announcing theme macrowords. Contrastiveness is a pragmatic overlay in the case of many focus constituents, especially presupposed information that is focused on. In other words, the three focus structures are used in certain strategies: 1. Predicate focus structures are used for commenting in topic-comment articulations. 2. Sentence focus structures are used for presentational sentences, and for themeredirecting and theme-supporting, event-reporting and state-reporting sentences. The word-order is generally marked. 3. Argument focus is used for unexpected, contrastive identification, and for the announcement of theme macrowords. The word-order is marked, similar to sentence focus structures. All the topic and focus categories and their respective information structure strategies have a link with the theme of a discourse. Theme has been defined in this study as the developing and coherent core or thread of a discourse in the mind of the speaker-author and hearerreader, functioning as the prominent macrostructure of the discourse (chapter 7 (7.4.4)). The information structure with its topics and focus structures and its strategies, can be used as a tool to identify and analyse themes. These categories and strategies together are called theme traces when they occur in marked syntactic constructions or in other prominence configurations like relexicalisation, end-weight, and repetition of macrowords. Theme traces are defined with the following wording: A theme trace is a clue in the surface form of a discourse, viewed from the perspective of information structure, that points to the cognitive macrostructure or theme of a text. This clue is in the form of (1) a marked syntactical configuration, be it marked word-order or marked in the sense of explicit and seemingly “redundant”, all signaling some thematic sequencing strategy, or (2) some recurring concept(s) signaling some prominence and coherence (chapter 7 (7.5.4)). By investigating these theme traces, the analyst will have a tool to study themes in discourse. This theme traces tool will assist in the demarcation of the sections in the developing theme of a text by means of a variety of boundary features, and once these thematic units have been established, the study of the topic framework together with the focus content will yield a verifiable understanding of the macrostructure of a text in Biblical Hebrew. Global themes are contrasted with local themes. Global themes occur in the higher-level thematic groupings, like whole narratives and smaller episodes within the narratives. Within the episodes are sub-units like scenes and thematic paragraphs, the smallest thematic unit. In scenes and thematic paragraphs, local themes occur. Between the different thematic units, a variety of theme sequential strategies occur. Theme shifting is a wider information structure strategy that is in operation in discourse. For instance, topic promotion, topic shift, and topic text-world framing are all cases of theme shifting. To study the theme of a narrative discourse from the perspective of the information structure, four steps of a theme-tracing model have been suggested, and applied to Genesis 17.

This dissertation concerns the dialogical encounters that take place between social workcr and client within the social services. The empirical material consists of 21 dialogues, 19 of which wcre recorded at two social welfare offices, one located in the center of Stockholm and the other in a suburb, and two which were recorded at the youth counseling office. In all, the material comprises 5150 individual utterances. In the literature, the type of dialogical cncounters studicd is called "institutional discourse". The theoretical approach in the analysis is based on an interactionist perspective, particularly with respect to dialogical analysis of verbal encounters. The aim of thc study is to analyse the dialogue between social worker and client from three points of de parture: participallt structure, topic, and inherent perspectives. In studying the discourse, the ini tiative-response (IR-) method of analysis was applied. The IR-method focuses on the dynamics of the dialogue, its pattern of dominance and its coherence. The results show that the social worker was interactively dominant in about two-thirds of the dialogues studied, but this dominance was most forcefully evident with respect to soliciting inititatives. With respect to content, the study concerns how the topical structure of the dialogue is constructed, i.e. which topics are taken up in the discourse and which of the parties steers the transition between topics. In the case of transitions to a topic defined as the main issue of the dialogue, it was the social worker who, in four-fifths of the cases, strategically steered the course of the dialogue. The perspective analysis undertaken in the study is concerned with investigating how different topics are treated in the dialogues, i.e. which of the parties "establishes" the perspectives and which "submits" to them . This analysis, made on a selection of seven of the 21 dialogical encounters, shows that it was the social worker who, in about two-thirds of thc cases, established by means of various interactive moves a bureaucratic or social welfare perspective on thc main topic sequences. Lastly, discursive dominance and power are discussed in the light of the prececling analyses.

Stories, themes, concepts and references are organized structurally and purposefully in most books. A person reading a book needs to understand themes and concepts within the context. Schank’s Dynamic Memory theory suggested that building on existing memory structures is essential to cognition and learning. Pirolli and Card emphasized the need to provide people with an independent and improved ability to access and understand information in their information seeking activities. Through a review of users’ reading behaviours and of existing e-Book user interfaces, we found that current e-Book browsers provide minimal support for comprehending the content of large and complex books. Readers of an e-Book need user interfaces that present and relate the organizational and narrative structures, and moreover, reveal the thematic structures. This thesis addresses the problem of providing readers with effective scaffolding of multiple structures of an e-Book in the user interface to support reading for comprehension. Recognising a story or topic as the basic unit in a book, we developed novel story segmentation techniques for discovering narrative segments, and adapted story linking techniques for linking narrative threads in semi-structured linear texts of an e-Book. We then designed an e-Book user interface to present the complex structures of the e-Book, as well as to assist the reader to discover these structures. We designed and developed evaluation methodologies to investigate reading and comprehension in e-Books, in order to assess the effectiveness of this user interface. We designed semi-directed reading tasks using a Story-Theme Map, and a set of corresponding measurements for the answers. We conducted user evaluations with book readers. Participants were asked to read stories, to browse and link related stories, and to identify major themes of stories in an e-Book. This thesis reports the experimental design and results in detail. The results confirmed that the e-Book interface helped readers perform reading tasks more effectively. The most important and interesting finding is that the interface proved to be more helpful to novice readers who had little background knowledge of the book. In addition, each component that supported the user interface was evaluated separately in a laboratory setting and, these results too are reported in the thesis.

Topic modeling refers to the process of algorithmically sorting documents into categories based on some common relationship between the documents. This common relationship between the documents is considered the “topic” of the documents. Sentiment analysis refers to the process of algorithmically sorting a document into a positive or negative category depending whether this document expresses a positive or negative opinion on its respective topic. In this paper, I consider the open problem of document classification into a topic category, as well as a sentiment category. This has a direct application to the retail industry where companies may want to scour the web in order to find documents (blogs, Amazon reviews, etc.) which both speak about their product, and give an opinion on their product (positive, negative or neutral). My solution to this problem uses a Non-negative Matrix Factorization (NMF) technique in order to determine the topic classifications of a document set, and further factors the matrix in order to discover the sentiment behind this category of product.

Syftet med denna studie har varit att undersöka hur olika ordföljdsprinciper i ett förstaspråk påverkar inlärningen av ordföljden i andraspråket svenska. Två grupper jämfördes där den ena hade ett förstaspråk med en bunden ordföljd baserad på den grammatiska principen och den andra ett förstaspråk med en fri ordföljd baserad på tema-rema-principen. Respondenterna bestod av elever inom den kommunala vuxenutbildningen i en västsvensk kommun som gick och läste på Sfi och SAS Grund. Varje respondent fick i uppgift att fritt skriva en kort uppsats. Texterna analyserades därefter ingående för att kunna identifiera vilka sats- och textgrammatiska ordföljdsfel som hade gjorts. Resultatet visade att de båda grupperna gjorde ungefär samma typ av satsgrammatiska fel men att gruppen med ett förstaspråk med fri ordföljd gjorde något fler textgrammatiska fel. Medan flera andra studier lyfter fram övergeneralisering av tema-rema-principen som det största problemet ur textgrammatisk synvinkel för denna senare grupp, fann jag att det snarare var underanvändning av tema-rema-principen som var problematiskt. / The purpose of this study was to examine how different word order principles in a first language influence the learning of word order in Swedish as second language. Two groups were compared in which one had a first language with a bound word order based on the grammatical principle and the other a first language with a free word order based on the topic-comment-principle. Respondents consisted of students in municipal adult education in a western Swedish municipality who studied at SFI and SAS Basic. Each respondent were asked to freely write a short essay. The texts were then analyzed in detail to identify which batch- and text grammatical word order inaccuracy that had been made. The results showed that both groups did about the same type of grammatical errors, but that the group with a first language with free word order did some more text grammatical errors. While several other studies highlight the generalization of the topic-comment-principle as the biggest problem for this latter group, I found that it was rather the lack of use of the topic-comment-principle that was problematic.

RB redactó el primer borrador del artículo y realizó la revisión final. EM redactó la versión presentada al comité editorial, elaboró las figuras del artículo, levantó las observaciones del comité editorial y redactó la versión final del artículo. AG realizó contribuciones al texto inicial, desarrolló la primera versión del resumen y revisó la versión final. GM realizó contribuciones al texto inicial, levantó las observaciones del comité editorial, realizó contribuciones al texto final y revisó la versión final. / Los estudios de síntesis (EDS): revisiones sistemáticas y metanálisis, son base para llevar a cabo las evaluaciones económicas en salud (EES). Los EDS, al permitir obtener parámetros para las probabilidades de estimación y de efectividad a partir de la combinación de resultados de estudios primarios y, al incluir en su metodología procesos de selección, evaluación, sistematización y síntesis, son considerados como el primer nivel de jerarquía en la evidencia científica. No obstante, pueden estar sometidos a sesgos y fallas metodológicas que afectan su validez. El presente artículo expone, inicialmente, la importancia de la aleatorización en la jerarquización de los diseños de investigación; luego se revisa los principales factores que afectan la validez de los EDS, incidiendo en el sesgo de publicación, la heterogeneidad y la inclusión de estudios primarios con objetivos principales distintos al del EDS; asimismo, se presenta a los estudios individuales como una alternativa válida para el desarrollo de una EES. Se concluye que uno de los aspectos claves en una EES es la selección adecuada de los tipos de estudio, sean estos primarios o secundarios. / Synthesis studies (SS): systematic review and meta-analysis are the basis for developing Health Economic Evaluations (HEE). SS allow us to obtain parameters for estimating probabilities and effectiveness from the combination of the results of primary studies, and, as they include in their methodology the selection, evaluation, systematization and synthesis processes, they are considered the first level of hierarchy in scientific evidence. Nevertheless, they can be prone to bias and methodological failures that can affect the validity of their results. This article initially presents the relevance of the randomization in the hierarchic classification of research designs, then it reviews the main factors affecting the validity of the SS, emphasising the publication bias, the heterogeneity and the inclusion of primary studies with main objective differing from the one of the SS. Moreover, it presents individual studies like a valid alternative for the development of a SS. The conclusion is that one of the key aspects in a SS is the correct evaluation of the study types and the objective evaluation of their quality, being these primary or secondary.

The main topic of the master's thesis is the phenomenon of Islamism in the Francophone Maghrebian literature. It focuses on the translation of novels from the region of Maghreb into the Czech language and on the reception of this type of literature in the Czech and French culture and society. The thesis is divided into two parts: theoretical and practical one. The main aim of the theoretical part is to be interdisciplinary; it explains the main concepts of the political science such as the concept of a nation or of immigration policies. Only a good understanding of these policies can lead to the full comprehension of Islamism in Western countries. Moreover, this part presents the main topics of translation studies, i.e. postcolonial studies, after the cultural turn and the main characteristics of the Maghrebian literature. The second part of the thesis deals with the reception of novels about Islamism in France and in the Czech Republic and analyses the Czech translation of a novel written by Boualem Sansal. Key words Arabic culture, Islamism, Francophone literature, Maghreb, translation, reception, Boualem Sansal

Doctor of Philosophy / Department of Computing and Information Sciences / William Henry Hsu / Topic models are probabilistic models for discovering topical themes in collections of documents. In real world applications, these models provide us with the means of organizing what would otherwise be unstructured collections. They can help us cluster a huge collection into different topics or find a subset of the collection that resembles the topical theme found in an article at hand. The first wave of topic models developed were able to discover the prevailing topics in a big collection of documents spanning a period of time. It was later realized that these time-invariant models were not capable of modeling 1) the time varying number of topics they discover and 2) the time changing structure of these topics. Few models were developed to address this two deficiencies. The online-hierarchical Dirichlet process models the documents with a time varying number of topics. It varies the structure of the topics over time as well. However, it relies on document order, not timestamps to evolve the model over time. The continuous-time dynamic topic model evolves topic structure in continuous-time. However, it uses a fixed number of topics over time. In this dissertation, I present a model, the continuous-time infinite dynamic topic model, that combines the advantages of these two models 1) the online-hierarchical Dirichlet process, and 2) the continuous-time dynamic topic model. More specifically, the model I present is a probabilistic topic model that does the following: 1) it changes the number of topics over continuous time, and 2) it changes the topic structure over continuous-time. I compared the model I developed with the two other models with different setting values. The results obtained were favorable to my model and showed the need for having a model that has a continuous-time varying number of topics and topic structure.