Data Products Final Project

Twitter Mining

Avinash Singh Pundhir
Analyst

Introduction

  1. This app will harvest tweets that have a matching string provided by user.

  2. A word cloud based on frequency of specific words will be provided.

  3. Following are some possible business usages of this application:

  • Sentiment analysis on twitter
  • Evaluate marketing initiative popularity
  • Learning and development of people new to R

User Interface

  1. User can fill in the string i.e. to be mined.

  2. User can select the number of top tweets they want to access using radio buttons.

  3. User can select total number of tweets to be pulled to get a meaningful word cloud using slider.

Proposed Enhancements

  1. Add a baysian sentiment analysis algorithm to identify associated sentiments.

  2. Enhance the total number of tweets limit that can be harvested.

  3. Add additional tools to allow users to mine twitter data in more detail from twitterR package ex. weekly, daily trends, specific tewwts from a particular user etc.

  4. Formatting data more accurately to remove duplicate entries and some mismatched searches for a more accurate analysis.

Basic Code Walkthrough

  1. Tweets mined for the word "data science""
top10
##  [1] "Check out a sample segment of Building Data Science Teams on"                                            
##  [2] "Listening totalk about data science Sexiest Job of the st Century for radiology Makin me happy imagingIT"
##  [3] "imagingITwebbinar on Data Science in Radiologyof the worlds data is onlyyears old"                       
##  [4] "DHISCOs blog reports on Big Data Challenge winnersUNT Computer Science and Engineerings Team SQL"        
##  [5] "Back by popular demand the FREE week DataScience amp MachineLearning EDX course Sign up"                 
##  [6] "Hellocould you please help me with the list of faculties for Master of Data Science program"             
##  [7] "sets standard for datascience reporting The Experts Were Wrong"                                          
##  [8] "New to Data Science Our day London workshop will get you up to speeddatascience"                         
##  [9] "RTTreat Customers Like Regulars with Data Science by billmanhvia CMSWire"                                
## [10] "Check out a sample segment of Building Data Science Teams on"