MS project desertation
Things Done :
-
Postgres connection.
-
Consumer Pool.
-
Trends from 8 different locations stored in DB date-wise.
-
Search urls built.
-
Stack of jobs completed.
-
Exectuor based approach.
-
Consumer Pool with 100 consumber objects instead of just app wise. This will help me fire threads.
-
Procedures to disintegrate the tweets
-
TFIDF computation
-
UI
-
Visualization
-
Build Trend - Entity Matrix
-
Understand what needs to go in as input into K-Medoid
-
Use k medoid to group trends -
-
Add sentiment analyzer
-
UI
-
Documentation
Twitter REST:
Getting Trends based on different locations without breaching rate limits.
Ability to add multiple accounts and maintain rate limit across all of them.
Database used: Postgres.