First, the list of all characters was acquired from Gilmore Girls fandom wiki using Selenium-Webdriver. Then, also using Seleninum, the scripts from each episodes were scraped from this website. The script dataset contains:
- character
- line
- season
Because the data didn't have any missing values, only small spelling mistakes (such as Loreai -> Lorelai) were manually corrected. Then using visualitaion tools (seaborn, matplotlib and NLP libraries (NLTK, re), the simple analysis of the Gilmore Girls script was conducted and placed into a dashboard. Moreover, a wordcloud was created.
There are multiple measures of centrality used in the network anlaysis. Four of them were tested and visually represented in this project:
- Degree Centrality:
- Betweenness Centrality:
- Closeness Centrality:
- Eigenvector Centrality: