/TweetsTrend

A Sentiment Analysis project for tweets using Spark, React, Spring Boot

Primary LanguageJavaScript

TweetsTrend

Created by Yuxiang Chen and Meihua Pan.

Step1: How to Deploy the Server Locally

  1. Setup mysql and configure database on your local machine

    • If you haven't install MySQL, install it! We recommend the MySQL5.7 Community Server
    • Open your terminal (If the password is not root, please reset it to root) 
      mysql -u <username> -p
# This will prompt you to enter your password
# default username and password are both: root
    • Create database called tweetsTrend 
      create database tweetsTrend;
use tweetsTrend;
    • Make sure the port for MySQL is 3306 (by default)

  2. Open the pom.xml in the server folder and install all maven dependencies for Spring Boot (If you open it with your IDE, it will automatically import all the dependencies usually).

  3. cd directory server/src/main/java/com.ai.tweetsTrend and run TweetsTrendApplication.java

  4. Once you see Started FormationHelperApplication in 4.277 seconds (JVM running for 5.134) in the console, the server side is successfully deployed locally.

Step2: How to Deploy Spark Locally and Stream Tweets from Twitter API

  1. Right now, we hard coded our own Twitter developer API (for any other purpose, please replace it with your own token)
  2. Pyspark runs on python2.7, it is not compatible with python3 for now.
  3. Currently, pyspark only support Java-jdk-8, other version will lead to bugs. Downgrade it if necessary.
  4. Install dependencies

    • pyspark : pip install pyspark

    • nltk : pip install nltk

    • vaderSentiment: pip install vaderSentiment

      1. Once you complete install vaderSentiment, please go in to the directory of vaderSentiment.
      2. Open vaderSentiment.py, at the top of the file, add 
        from io import open

    • tweepy: pip install tweepy

  5. open two separate terminals, in both terminals: cd to the project directory (../TweetsTrend)
  6. Inside the first terminal, run 
    python tweets.py
  7. Inside the second terminal, run 
    python spark.py

Step3: How to Open the Web Page Locally

  1. Make sure your computer has the latest node.js installed
  2. open a terminal
  3. cd into the current project directory ../TweetsTrend 
    cd web
npm install
npm start
  4. Once it starts running, the website will automatically open in your default browser with address localhost:3000/