How otniel113/hebubaninfluencer works
Project type and purpose
• This is a two‐phase research pipeline (data‐gathering + data‐analysis) rather than a full‐stack or microservices system.
• Purpose: crawl tweets on the game “Heaven Burns Red,” build a social‐network graph, compute Eigenvector Centrality, and list top influencers.
File structure
• Top level contains:
• 01.CrawlingHebuban.py – the standalone Python crawler that calls the Twitter API.
• 02.ModelHebuban.ipynb – Jupyter notebook to load stored data, build the graph (NetworkX), compute centrality, and generate visualizations.
• README.md – explains usage, dependencies, and research context.
README for technical stack & patterns
• External service: Twitter API (requires a bearer token).
• Libraries and frameworks: Python 3.x, Tweepy or requests, pandas, NetworkX, Jupyter.
• Data format: JSON or CSV intermediate (tweets, user IDs).
a. External Twitter API (external service)
b. Crawler module (01.CrawlingHebuban.py)
c. Local data store (raw tweet data in JSON/CSV)
d. Analysis module (02.ModelHebuban.ipynb running pandas + NetworkX)
e. Visualization/report outputs (plots, tables of influencers)
f. Operator/User (launches the crawler, opens the notebook)
Architectural patterns & design principles
• Clear separation of concerns: crawling code vs. analysis code.
• Single‐node deployment: all steps run locally (or on a single VM).
Relevant technologies/frameworks
• Python script dependencies: requests or Tweepy, JSON, CSV, logging.
• Notebook dependencies: pandas, NetworkX, matplotlib or seaborn.
• External: Twitter Developer API (Bearer Token).
• Execution environments: CLI for crawler; Jupyter for analysis.