UCLA

This thesis primarily employs text mining techniques to analyze the novel series ”Remem- brance of Earth’s Past” by Liu Cixin, with the aim of extracting information and develop- ing a comprehensive understanding through this process. It involves extracting the most frequently occurring words, calculating tf-idf (term frequency-inverse document frequency) scores, validating Zipf’s Law, generating bi-grams, conducting sentiment analysis and apply- ing topic modeling. The findings reveal significant keywords from the novel series, allowing readers to infer the main elements. While this analysis can help the audience achieve an overall understanding of the plots, reading the original work is encouraged to have a detailed sense of the entire series.