Algoritmer är ett växande fenomen som blivit allt mer involverat i våra liv. Personer mellan18-30 år använder sociala medier i störst utsträckning och är den generation som anser sigmest kunnig om algoritmer. Tidigare studier påvisar användare av sociala medier endast harbegränsad kunskap rörande algoritmer och postulerar att forskning behöver göras för attfastställa konsekvenserna av att använda algoritmer på sociala medier. Med ökad användningav algoritmer på sociala plattformar är det av intresse att undersöka den åldersgrupp somanvänder sociala medier i störst utsträckning. Syftet med studien är att undersöka sambandetmellan kunskap, attityd och beteende gentemot algoritmbaserad anpassning på socialamedier. Det undersöktes genom en kvantitativ enkätundersökning med teoretisk grund imodellen Knowledge-Attitude-Behavior (KAB). Resultatet visade positiva samband mellandessa vilket kan sammanfattas som att användare med hög kunskap om anpassning har enpositiv attityd till konceptet relevansbaserad anpassning och anser att informationen omsådan är tillgänglig. Användares beteende visar på att de använder sig av plattformars verktygför att påverka anpassning själv och att det påverkar vilka plattformar de väljer att använda,samt huruvida man är villig att börja bruka en ny plattform. Denna studie har även identifieratatt KAB är ett användbart ramverk att tillämpa för denna typen av studier och lärdomar förframtida forskning framförs. / Algorithms are a growing phenomenon that has become increasingly involved in our lives.People aged 18-30 use social media the most and are the generation who consider themselvesthe most knowledgeable about algorithms. Previous studies demonstrate that social mediausers have only limited knowledge regarding algorithms and postulate that research needs tobe done to determine the implications of using algorithms on social media. With the increaseduse of algorithms on social platforms, it is of interest to examine the age group that usessocial media the most. The purpose of this study is to investigate the relationship betweenknowledge, attitude and behaviour towards algorithm-based curation on social media. It wasinvestigated through a quantitative survey with a theoretical basis in theKnowledge-Attitude-Behavior (KAB) model. The results showed positive correlationsbetween these, which can be summarized as users with high knowledge about curation have apositive attitude towards the concept of relevance-based curation and are of the opinion thatinformation about it is available. Such users' behaviour indicates that they use platform toolsto influence customisation themselves and that this influences which platforms they choose touse, as well as whether they are willing to start using a new platform. This study has alsoidentified that KAB is a useful framework to apply for this type of study and insights gainedare presented for future research.

Denna uppsats syftar till att undersöka hur filterbubblor och ekokammare skapar politisk polarisering. För att uppnå detta syfte appliceras en diskursanalys på debatten om koranbränning. Trådar på Flashback forum och kommentarsfält på Twitter undersöks för att urskilja eventuella retoriska strategier, diskursiva yttringar och sociokulturella aspekter som kan bidra till polarisering. För att uppnå detta appliceras Fairclough’s tredimensionella modell för att utföra en kritisk diskursanalys av innehållet på de definierade filterbubblorna / ekokammarna. / The aim of this study is to analyze how filterbubbles and echochambers affects political polarization. A discourse analysis will be applied on the debate that has broken out due to recent burnings of Qurans. More specifically, this method will be applied on Twitter and Flashback to study rhetorical strategies, discursive utterances and sociocultural views that is affecting polarization. Fairclough’s three-dimensional model will be applied to perform a critical discourse analysis on the defined filterbubbles / echochambers.

This thesis is about the phenomenon filter bubble and how it affects the participant democracy. The purpose of this thesis is to analyze how participant democracy is presented together with the filter bubbles in articles of filter bubbles. The questions that is answered inthis thesis is: In which way is filter bubble described in the Swedish newspapers “Dagens Nyheter” and “Svenska Dagbladet”? How is participant democracy presented in relation tofilter bubbles? The results show us that the Swedish newspapers DN and SvD is presenting the filter bubbles in their articles and has different arguments that goes within the participant democracy.

Today, it has become a natural part of life for many people to share personal information and turn to search engines and social media as they search and seek information on the internet. This study aims to investigate how individuals are affected by Big Data in their interactions with social media and to understand how individuals relate to personalization of Big Data, digital identity, agency and how they share their own personal data. This has been researched through a case study involving qualitative interviews and experiments. The result of the research shows that personalization of Big Data on social media has an impact on individuals in their ever yday lives. The research from this case study shows a certain negative attitude towards personalization on social media and that there is so me awareness of the companies agency on the internet, this does not affect individuals in taking any actions. This leads to the conclusion that personalization is to some extent a desired function. / Idag har det blivit en självklarhet för många att dela med sig av personlig information på internet och vända sig till sökmotorer och sociala medier när de letar och söker information. Studien syftar till att undersöka hur individer påverkas av Big Data i sin interaktion med sociala medier och att förstå hur individer förhåller sig till personalisering av Big data, digital identitet, aktörers makt och hur de tänker kring delning av personlig data. Detta studeras genom en fallstudie där kvalitativa intervjuer och experiment. Resultatet från studien visar att personalisering av Big Data på sociala medier påverkar individer och deras digitala användning i vardagen. Studien visar en viss negativ inställning till personalisering på sociala medier och på viss medvetenhet kring aktörernas makt på internet, detta påverkar inte individerna att göra aktiva handlingar. Detta leder till slutsatsen att personalisering av Big Data i viss mån är en accepterad funktion, även om det innebär en förlorad makt över sin digitala användning.

Participatory culture inside a filter bubble – A content analysis of the network society Twitch.tv’s external communication. The aim of this essay was to study how user-generated content is being used and how it is presented by Twitch Interactive, Inc. in its external communication, more specifically how the discourse of this content is shaped on Twitter. To gather the information needed a quantity based content analysis was made, ranging from 1-31st of October 2015. From this data patterns where detected and a discourse analysis based on Berglez version of van Dijks model was conducted on six of the Twitter posts that followed said pattern. This was done to answer how the discourse was shaped and what the consequences could be of using such a pattern. Analyzing the results was done using theory from strategic marketing researchers such as Larsåke Larsson, Lars Palm, Ulf Dahlqvist and Magnus Linde. Furthermore theorist like Gripsrud, Pariser, Lindgren, Jenkins, Ford and Green were used to describe and analyze the field of culture, more specifically participatory and popular culture. The conclusion of the study was that Twitch representation of user-generated material and the culture it represents was unjustified and could be seen as a filter bubble. The results showed that the majority of content was young white male gamers where as other research on the field of gaming culture shows a more equal split between genders. The individuals that where represented in the content could be seen as objects of admiration and association. Discussion led to the belief that Twitch may not possess the power to change to discourse due to the strong economic need of the consumers. This need was concluded to arise from the fact that the platform and its culture is much like participatory culture and where this culture has been taken to a commercial level.