Watson-TweetlyzR

Rudy Martin - Developing Data Products
Sept2016

How to Gain Friends and Still Have Time for Yourself

Quality Relationships Require More Time, Skill

  1. Constant stream of online chatter 24/7 on Twitter
  2. Not all online personas are worth interacting with
  3. Some players are even dangerous
  4. Communicating online requires new speaking muscles

Solution: Our Social Matching Recipe

  1. Add One Part IBM Watson Personality Profile
  2. Add One Part Social Media Text - Extracted Tweets
  3. Mix-in with R Data Mining and Exploratory Analysis
  4. Serve Online via Interactive Shiny Visual Tool

Secret Ingredients: Dynamic Data and Radarchart

chartJSRadar(scores[, c("Label", input$selectedPersonas)], 
    maxScale = maxScaleR(),
    scaleStepWidth = scaleStepWidthR(),
    scaleStartValue = input$scaleStartValue,
    responsive = input$responsive,
    labelSize = input$labelSize,
    addDots = input$addDots,
    lineAlpha = input$lineAlpha,
    polyAlpha = input$polyAlpha,
    showToolTipLabel=input$showToolTipLabel,
    colMatrix = v$cmText)

Watson-TweetlyzR : Fun Social Research

Caption for the picture.

Next Steps

  1. Download Github code and sample data

    App: https://rudymartin.shinyapps.io/Watson-TweetlyzR/

    Source code: https://github.com/RudyMartin/Watson-TweetlyzR

  2. Signup for IBM/Twitter developer accounts,insert credentials

  3. Insert your credentials and run in RStudio

  4. Have fun seeing what your friends are really saying

  5. Be a friend (Twitter:@realrudymartin Email: realrudymartin@gmail.com)