Sentiment mining in Goodreads reviews of classic American novels
This bachelor’s thesis analyses the sentiment used in 1 and 5 star Goodreads reviews for 10 classic American novels. The sentiment analysis is done with the SentiStrength programme and the results are reviewed and compared to review lengths and average Goodreads ratings. This is done to answer two research questions: 1. Is more sentiment expressed in 1 or 5 star reviews? 2. Which of the chosen books has the highest sentiment in its reviews? The thesis is divided into five sections: the introductions, theoretical background, overview of the research method, analysis of the sentiment scores and conclusion. The introduction discusses why the topic was chosen. The theoretical part is divided into four sections. The first gives an outline of the democratisation of expertise in different fields, including the humanities. The second part talks about digital humanities and distant reading. The third section talks about sentiment analysis the fourth introduces the SentiStrength programme. The theoretical part is followed by an outline of the method, the research findings and a discussion of the findings. The thesis then discusses its limitations and ideas for future research.
Die folgenden Lizenzbestimmungen sind mit dieser Ressource verbunden: