TwittMap-sentiment-server

COMSE6998 CLOUD COMPUTING & BIG DATA - TwittMap(Assignment2)

Team Member

Fan Su (fs2488) Jingyi Guo (jg3421) Wei Cao (wc2467) Shijie Hu (sh3251)

website for demo

http://twittapp-env.elasticbeanstalk.com/index.jsp

Files

README - this file

Internal Design

Features:

1. Display twitts' location on map with predefined keywords search ('halloween', 'USAirway', 'NewYork', 'Columbia')
2. Display twitts' location based on time range filter
3. Color gradient and density map display for twitts on map based on location
4. Real time twitts location display
5. Auto create AWS Elastic beanstalk application, environment, and deploy TwittMap application
6. Sentiment Analysis for text acquired from Twitt API, and real-time display how many positve, neutral, negative text up to now
7. Worker in acquiring data from queue, and analyze text

Tools used:

1. Web server: Tomcat 7.0 on AWS Elastic Beanstalk
2. Database: Apache Cassandra on AWS EC2
3. API: Twitter Live and Streaming API, Google Map API,Elastic Beanstalk,Elastic LoadBalancing
4. AWS SQS, SNS, and sentiment analysis API alchemyapi

Steps to use without installing source code:

1. Go to page http://twittapp-env.elasticbeanstalk.com/
2. 'Main' will show main feature with locations of twitts, keywords filtering in dropdown, time range filtering
3. 'Real time' will show real-time twitts' location with predefined search keywords 'halloween'

Steps in using source code

1. Git clone source code
2. Link to AWS account with access key and secret key and connect with current development environment
3. Create application and environment on AWS Elastic Beanstalk
4. Add jars to build path
5. Create Cassandra Database on AWS EC2
6. Create account in Twitter API and get access keys
7. Deploy project in AWS Elastic Beanstalk

Comment to TA

If there is anything wrong, please contact us!