/Sentiment-Analysis-of-Tweets

Primary LanguageJupyter NotebookApache License 2.0Apache-2.0

Sentiment-Analysis-of-Tweets

Explores and runs a sentiment analysis using VADER approach on about 25k tweets mentioning @Dell to label each tweet as either positive, negative or neutral.

VADER (Valence Aware Dictionary and sEntiment Reasoner)

  • Uses "bag of words" approach
    1. Removes stop words
    2. each word is scored and combined to give a total score
    3. However, this model does not account for the relations between texts

Steps & Modules:

  • pandas - Performs EDA
  • matplotlib and seaborn - visualization
  • nltk’s SentimentIntensityAnalyzer- - runs VADER model
  • Labels tweets as positive, negative or neutral
  • wordcloud - to build wordcloud
  • Interesting insights from further EDA

This is work in progress notebook, more analysis to come! Thanks