/zgeorgiou-ADE2017-Characterizing-Popularity-Virality-Sentimentalism-of-videos-on-YouTube-and-Twitter

Primary LanguageJavaMIT LicenseMIT

Yttresearch

This project contains all the tools developed for the purposes of my bachelor thesis on Characterizing the Popularity, Virality and Sentimentalism of Video Content Categories on YouTube and Twitter

  • Collect YouTube & Twitter data
  • Analyze the data

Analysis on how YouTube videos disseminate through social media

This project is divided into 2 main/core services:

  1. Collector
  2. Analyzer

1. Video & Tweets Collector

Features

TPS: Twitter Public Stream

  • Finds new videos from TPS and stores their static data into db.
  • Collects tweets that mention YouTube videos through TPS.
  • Collects the dynamic information of YouTube videos once per day. Currently is cummulative
  • Collects video comments.
  • Removes incomplete Videos video comments.
  • All data is stored into the following collections:
    • Videos
    • Tweets
    • Comments
    • YouTubeKeys
    • TwitterApps
    • configurations
  • Commands are handled via Rest API - Authentication Token Needed

2. Analyzer

Features

  • Sentiment Analysis
  • Statistical Analysis
  • Videos Information
  • Groups Information
  • Classification
  • Feature Importance Analysis