TY - THES A1 - Nolte, Johannes T1 - Untersuchung der Vorgehensweise ausgewählter Internetplattformen gegen gefälschte Online-Bewertungen N2 - Diese Arbeit dokumentiert ein Experiment zur Untersuchung des Umgangs von Facebook, Google Maps, Google Play Store und Zalando mit gefälschten Bewertungen („Opinion Spam“). Es wurde untersucht, ob die Plattformen im Rahmen dieser Arbeit abgegebene gefälschte Bewertungen veröffentlichen. Dazu wurden je Plattform 24 unterschiedlich authentisch wirkende Konten erstellt. Mit diesen wurden je 48 gefälschte Bewertungen nach folgenden Kriterien abgegeben: 1) Kopierte Produkt-/Ortsbeschreibung oder Werbesprache 2) Aussagen ohne Produkt-/Ortsbezug oder inhaltliche Fehler 3) Blindtexte oder sehr viele Rechtschreibfehler 4) eine Meinung verknüpft mit der Aussage, dass der Bewertende das Produkt/den Ort nicht kennt. Innerhalb von 14 Tagen wurde mehrmals überprüft, ob eine Bewertung freigeschaltet oder gelöscht wurde. Bei Facebook wurden 6,7% der abgegebenen Bewertungen innerhalb der ersten 14 Tage gelöscht. Nachdem die verbleibenden Bewertungen der Plattform gemeldet wurden, waren es 47,7%. Bei Google Maps wurden 4,2% entfernt, im Google Play Store 12,5% (bei beiden keine Änderung durch Melden). Bei Zalando (kein Melden möglich) war die eigenständige Löschquote mit 43,8% mit Abstand am höchsten. Besonders häufig wurden im Experiment Blindtexte gelöscht. Positive und negative Bewertungen wurden nahezu gleich oft entfernt. Bewertungen von Konten mit simulierter Nutzerinteraktion wurden häufiger gelöscht als Bewertungen von Konten ohne Nutzerinteraktion. Das Experiment hat gezeigt, dass sehr viele gefälschte Bewertungen, die eindeutig den Richtlinien der Plattformen widersprechen, nicht gelöscht wurden. Deren Sichtbarkeit variiert allerdings je nach Plattform und Produkt/Ort stark. Der Aufwand für die Erstellung vieler Konten bei den Diensten ist überschaubar. Wenn Nutzer gefälschte Bewertungen nach den im Experiment untersuchten Kriterien abgeben wollen, werden sie von den Plattformen daran wenig gehindert. N2 - This bachelor thesis contains the documentation of an experiment to evaluate how Facebook, Google Maps, Google Play Store and Zalando deal with fake reviews (“opinion spam”). Measured was whether the platforms publish or delete fake reviews submitted in the context of this work. For this purpose, 24 accounts with different attributes according to age and user interaction were created on each platform. With those accounts 48 reviews per platform were submitted based on the following criteria: 1) Copied product/location description or advertising-language 2) reviews without product/location reference or contextual errors 3) Blind texts or many spelling mistakes 4) reviews which contain an opinion but state that the author does not know the product/location. 6.7% of the reviews submitted to Facebook were deleted within the first 14 days, 47.7% after reporting the remaining reviews to Facebook. Google Maps deleted 4.2%, Google Play Store 12.5% (both showed no changes after reporting). Zalando (no reporting possible) has the by far highest independent deletion rate with 43.8%. Blind texts were most often deleted during the experiment. Positive and negative ratings were deleted almost the same number of times. Reviews from accounts with simulated user interaction were deleted more often than reviews from accounts without user interaction. The experiment showed that many fake reviews, which contradict the guidelines of the platforms, were not deleted. But their visibility varies depending on platform and product/location. The creation of many accounts at the services does not require too much effort. If users want to publish fake reviews according to the criteria used in the experiment, the platforms do not really hinder them. KW - Gefälschte Bewertungen KW - Opinion spam KW - Fake reviews Y1 - 2018 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:hbz:1383-opus4-2126 ER - TY - THES A1 - Oyewole, Oluwabusayomi T1 - Conversational design: the use of chatbot as an electoral information dissemination tool N2 - A steady decline in voter turnout has been recorded in many democratic countries, with young people (i.e Gen Z and Millennials) identified as a significant portion of those who don’t vote. Research has shown that one of the reasons young people do not vote is a lack of information about the electoral process, candidates and policies. Studies have also shown that preferred information dissemination medium is aligned across age groups, with older people trusting old media such as newspapers and young people leaning towards new media such as social media. With the recent advancement in machine learning, natural language interface and popularity of conversational agents such as chatbots, a new subset of new media has emerged. This study aims to investigate if chatbots could be an efficient medium for young people to access electoral information. It does this by comparing it with a website, using the EU electoral page as a case study. Based on the content of the EU electoral webpage, a chatbot was designed, trained and deployed on Telegram. Participants were randomly divided into two experimental groups (chatbot and website) and asked to find a list of crucial information, interact with the medium and then respond to the System usability scale (SUS), and a perception questionnaire. Observed data and interview notes were also taken. Analysis of the experimental data, observed data and surveys demonstrated that chatbot as a tool is inefficient as an electoral information tool because of the limitation of chatbot when out of scope questions are asked and limitation in following conversation contexts which is crucial to conversations. However, it is rated as above average on usability and joy of use, but raises questions on trust and data protection. It is recommended that the study be performed again when there are more advances to natural language processing to see if its efficiency as an electoral information medium will improve. A longitudinal study is also recommended to see if there’s a direct impact of chatbots as a tool, on young people’s motivation to vote. KW - Conversational design KW - Information KW - Chatbot KW - Election Y1 - 2020 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:hbz:1383-opus4-4294 ER - TY - THES A1 - Zelle, Philipp T1 - Untersuchung der npm Registry auf die Codequalität ihrer Pakete N2 - The aim of this project was to check whether the code quality of a npm package has an impact on some of its other properties. To check possible connections, lots of data had to be gathered. This was done by collecting available information from the GitHub and npm API about each package. Additionally, packages were downloaded to analyze them with cloc for data about the amount of lines of code, and with eslint to check the package for possible errors. This process was conducted for almost all packages with a public GitHub repository, and in more detail for the most downloaded packages. Results show that for top packages less errors lead to more popularity, this does not apply to the whole npm registry though. Furthermore, in both cases, more con- tributors in the repository lead to fewer errors. Lastly, it’s shown that the more recent packages were updated, the more popular they were and the count of packages in- creases the more recent their update was. Y1 - 2022 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:hbz:1383-opus4-13148 ER -