This project crawls financial tweets from specified accounts that post financial tweets and builds a datasets than can be used to perform sentiment analysis on any financial tweet that has financial tag. The dataset set has been used in performing sentiment analysis on tweets that bear companies listed on NSE.
Procedures carried out:
-
Crawling dataset from twitter. Picked out a few acoount mostly Kenyan that post news with emphasise on finacial news.
-
Cleaning up the dataset
-
labeling the dataset
-
Training and testing the dataset on the labelled dataset.
-
Testing dataset on unlabelled dataset to get the sentiment value.