epflexclt / ntds2019_team9

Aggregation Phenomena in Movie Industry

Team Members: Rui CHEN, Shichao JIA, Jimin WANG, Zhuoyue WANG

Requirements

• Python 3
• pandas
• numpy
• networkx 2.3
• pyecharts
• json
• datetime
• math
• seaborn
• matplotlib
• snapshot_selenium
• wordcloud
• nltk
• io
• PIL
• itertools
• scipy
• collections

Dataset

• tmdb_5000_movies.csv
  • General data of 4803 movies with 20 feature columns
    • budget
    • genres
    • homepage
    • id
    • keywords
    • original_language
    • original_title
    • overview
    • popularity
    • production_companies
    • production_countries
    • release_date
    • revenue
    • runtime
    • spoken_languages
    • status
    • tagline
    • title
    • vote_average
    • vote_count
• tmdb_5000_credits.csv
  • Credits data of 4803 movies with 4 feature columns
    • movie_id
    • title
    • cast
    • crew

Jupyter notebook

• ntds_tmdb_data_explore.ipynb

  • Preprocessing of original data, mainly including data cleaning and feature augmentation.
  • Preliminary data exploration

• ntds_tmdb_movie_graph.ipynb

  • Pipeline of Movie Graph built by similarity of leading actors, including data preprocessing, network construction, data analysis and visualization.

• ntds_tmdb_actor_graph.ipynb

  • Pipeline of Actor Graph built by co-appearance, including data preprocessing, network construction, data analysis and visualization.

• ntds_tmdb_crew_graph.ipynb

  • Pipeline of Movie Graph built by similarity of main crew, including data preprocessing, network construction, data analysis and visualization.

Component

• chromedriver.exe
  • This component is to ensure pyecharts work normally in jupyter lab by chrome.