5 research outputs found
Användning av sentimentanalys för att analysera ändringarna av allmänhetens uppfattning om politiska händelser
No full textSentiment analysis is a valuable tool for understanding and evaluating the tone and content of documents, articles and text. It can be used in conjunction with social media, to explore how public opinion can change over time, with a focus on political events. In this thesis, sentiment analysis together with clustering has been used to gain insight into that subject, using the 2024 Romanian presidential election as a relevant case study. By tracking changes in people´s perceptions of events over time, the effect social media can have on political events is analyzed in more detail. To achieve this, two sentiment analysis models were developed using the BERT and FastText libraries, and these models were trained to support both Romanian and English. After sentiment analysis, clustering was performed to further divide the results into distinct topics. This was achieved by using the spectral clustering method together with K-means. With a total of 8 494 tweets collected from the social media platform X before and after the first round of the election, a distinct negative shift could be observed, indicating a general dissatisfaction and a shift of the discussion regarding the election. Sentiment analysis was successfully performed, resulting in observable changes between before and after the election. Clustering helped provide more precise insight into which topics the changes in sentiment occurred. However, due to limited data before the election, especially in English, only the Romanian tweets could be clustered and the reliability is difficult to confirm. Further work is needed to be able to draw more general statements about political events, including collecting more data from various social media platforms and increasing the number of case studies to see if similar patterns can be observed.Sentimentanalys är ett viktigt verktyg för att förstå och utvärdera ton och innehåll i dokument, artiklar och texter. Det kan användas tillsammans med sociala medier för att undersöka hur den allmänna uppfattningen kan förändras över tid, med fokus på politiska händelser. I detta arbete har sentimentanalys tillsammans med klusteranalys använts för att få insikt i detta ämne, med det rumänska presidentvalet 2024 som en relevant fallstudie. Genom att spåra förändringar i människors uppfattning om händelser över tid, analyseras den påverkan som sociala medier kan ha på politiska händelser i mer detalj. För att uppnå detta utvecklades två sentimentanalysmodeller med hjälp av biblioteken BERT och FastText, och dessa modeller tränades för att stödja både rumänska och engelska. Efter sentimentanalysen utfördes klusteranalys för att ytterligare dela upp resultaten i distinkta ämnen. Detta uppnåddes genom att använda spectral clustering i kombination med K-means. Med totalt 8 494 tweets insamlade från den social media plattformen X före och efter den första valomgången, var det möjligt att observera en tydlig negativt ändring, vilket indikerar en allmän missnöjdhet och förändring i diskussionen kring valet. Sentimentanalysen genomfördes framgångsrikt och resulterade i observerbara förändringar mellan före och efter valet. Klusteranalysen hjälpte till att ge en mer precis insikt i vilka ämnen förändringarna i sentimentet skede. Dock, på grund av begränsad data före valet, särskilt på engelska, kunde endast de rumänska tweetsen grupperas, vilket gör att tillförlitligheten är svår att bekräfta. Ytterligare arbete krävs för att kunna dra mer generella slutsatser om politiska händelser, vilket inkluderar bland annat att samla in mer data från olika sociala medieplattformar och att öka antalet fallstudier, för att se om det går att observera liknande mönster
Användning av sentimentanalys för att analysera ändringarna av allmänhetens uppfattning om politiska händelser
No full textSentiment analysis is a valuable tool for understanding and evaluating the tone and content of documents, articles and text. It can be used in conjunction with social media, to explore how public opinion can change over time, with a focus on political events. In this thesis, sentiment analysis together with clustering has been used to gain insight into that subject, using the 2024 Romanian presidential election as a relevant case study. By tracking changes in people´s perceptions of events over time, the effect social media can have on political events is analyzed in more detail. To achieve this, two sentiment analysis models were developed using the BERT and FastText libraries, and these models were trained to support both Romanian and English. After sentiment analysis, clustering was performed to further divide the results into distinct topics. This was achieved by using the spectral clustering method together with K-means. With a total of 8 494 tweets collected from the social media platform X before and after the first round of the election, a distinct negative shift could be observed, indicating a general dissatisfaction and a shift of the discussion regarding the election. Sentiment analysis was successfully performed, resulting in observable changes between before and after the election. Clustering helped provide more precise insight into which topics the changes in sentiment occurred. However, due to limited data before the election, especially in English, only the Romanian tweets could be clustered and the reliability is difficult to confirm. Further work is needed to be able to draw more general statements about political events, including collecting more data from various social media platforms and increasing the number of case studies to see if similar patterns can be observed.Sentimentanalys är ett viktigt verktyg för att förstå och utvärdera ton och innehåll i dokument, artiklar och texter. Det kan användas tillsammans med sociala medier för att undersöka hur den allmänna uppfattningen kan förändras över tid, med fokus på politiska händelser. I detta arbete har sentimentanalys tillsammans med klusteranalys använts för att få insikt i detta ämne, med det rumänska presidentvalet 2024 som en relevant fallstudie. Genom att spåra förändringar i människors uppfattning om händelser över tid, analyseras den påverkan som sociala medier kan ha på politiska händelser i mer detalj. För att uppnå detta utvecklades två sentimentanalysmodeller med hjälp av biblioteken BERT och FastText, och dessa modeller tränades för att stödja både rumänska och engelska. Efter sentimentanalysen utfördes klusteranalys för att ytterligare dela upp resultaten i distinkta ämnen. Detta uppnåddes genom att använda spectral clustering i kombination med K-means. Med totalt 8 494 tweets insamlade från den social media plattformen X före och efter den första valomgången, var det möjligt att observera en tydlig negativt ändring, vilket indikerar en allmän missnöjdhet och förändring i diskussionen kring valet. Sentimentanalysen genomfördes framgångsrikt och resulterade i observerbara förändringar mellan före och efter valet. Klusteranalysen hjälpte till att ge en mer precis insikt i vilka ämnen förändringarna i sentimentet skede. Dock, på grund av begränsad data före valet, särskilt på engelska, kunde endast de rumänska tweetsen grupperas, vilket gör att tillförlitligheten är svår att bekräfta. Ytterligare arbete krävs för att kunna dra mer generella slutsatser om politiska händelser, vilket inkluderar bland annat att samla in mer data från olika sociala medieplattformar och att öka antalet fallstudier, för att se om det går att observera liknande mönster
Power, value, and the individual exchange: towards an improved conceptualization of terrorist finance
No full textThis thesis finds that the term ‘terrorist financing’ is a misnomer in that much of the activity encompassed by that term involves neither terrorism nor money. Instead, terrorist financing more accurately refers either to the flow of economic and material value to ‘terrorist’ actors or specific material expressions of support to ‘terrorism,’ however that contested term is defined. This finding not only directly challenges the dominant ways terrorist finance is now conceptualized, but also provides the first unified coherent conceptual framework capable of supporting systematic analysis of the topic. This thesis arrives at this conclusion by first critically examining the various – and often contradictory or incoherent – normative, legal, and political contexts that dominate ‘orthodox’ thinking on terrorism and terrorist finance, and then relocating the financing of terrorism squarely in context of the everyday realities of how terrorism and terrorist actors interact with global and local political economies. This thesis goes beyond existing critical works on terrorist financing, and constructs the necessary conceptual foundation for a vastly more coherent, systematic, and ultimately useful understanding of the financial and economic dimensions of terrorism
Making vision into power : Britain's acquisition of the world's first radar-based integrated air defence system 1935 - 1941
No full textThis thesis represents the first application of a current conceptual model of defence acquisition to analyse the historical process, the 1935 - 1941 British acquisition of an integrated air defence system pivoted upon the innovative technology of radar. For successful acquisition of a military capability, the model posits that balanced attention must be focused acoss eight 'lines of developmen' - not only equipment, but also doctrine and concepts, logistics, structures, personnel, organisation, training and information with an overarching requirement for interoperability. This thesis contrasts what turned out to be a successful acquisition, of radar to achive air interception capability by day in the Battle of Britain, with less successful acquisition, or radar to achieve the same capability at night, where an effective system arrived too late to ward off the Blitz. The results establish the validity of the model and its attendant lines of development concepts, and furnish new insights into acquisition processes and military history. Acquisition lessons are derived for the capability-based involvement of industry, for the experience and personality necessary for key managers at different 'life stages' of an acquisition and for the avoidance of over-rapid 'dysfunctional diffusion' of innovative technologies. Historical insights for the Battle of Britain include the sub-optimal performance, for trivial reasons, of key South Coast radars, and the critical importance of the human elements of the radar-based air defence system. For the Blitz, airborne radar hardware has previously been identified as a key problem, whereas research here exposes the greater need for accurate ground control radar, the sound selection and training of pilots and operators in new tactics, and provision of equipment maintainers and test gear. New evidence illustrates that pursuit of an alternative to radar significantly delayed the optimal solution, and throws fresh light both on personalities and on development process management
Nuclear EGFR modulation of DNA repair
No full textOverexpression of the epidermal growth factor receptor (EGFR) is associated with
resistance to chemotherapy and radiotherapy. EGFR involvement, in repair of radiation-induced
DNA damage, is mediated by association with the catalytic subunit of DNA
protein kinase (DNAPKcs). This study investigated the role of EGFR nuclear import,
and its association with DNAPKcs, on DNA repair following treatment either with
cisplatin or ionizing radiation (IR). EGFR- null murine NIH3T3 cells were transfected
with wild type or with mutated EGFR (mutations found in human cancers L858R,
EGFRvIII and mutations in the EGFR nuclear localization signal (NLS) sequence
NLS123, LNLS123). Comet assay analysis, which measures unhooking of cisplatin
crosslinks and repair of IR induced strand breaks, demonstrated that wtEGFR and
EGFRvIII completely repair cisplatin and IR induced DNA damage.
Immunoprecipitation studies show that repair is associated with the binding of both
wtEGFR and EGFRvIII to DNAPKcs, which increases by 2- fold 18 hours following
cisplatin treatment. Confocal analysis and proximity ligation assay indicated that this
association takes place both in the cytoplasm and in the nucleus resulting in a significant
increase of DNA-PK kinase activity. Intermediate levels of repair as shown by the
L858R construct with impaired nuclear localization demonstrated that EGFR kinase
activity is partially involved in repair but is not sufficient to determine EGFR nuclear
expression. EGFR-NLS mutants showed impaired nuclear localization and impaired
DNAPKcs association resulting in significant inhibition of DNA repair and downregulation
of DNA-PK kinase activity.
Our data suggest that EGFR nuclear localization is required for the modulation of
cisplatin and IR induced DNA damage repair. The EGFR-DNAPKcs binding is
triggered by cisplatin or IR and not by EGFR nuclear translocation per se.
Understanding mechanisms regulating EGFR subcellular distribution in relation to DNA
repair kinetics will be a critical determinant of improved molecular targeting and
response to therapy
