Recently I created a script pulling altcoin related twitter data. Now I have added some code to do textual sentiment analysis on the Twitter data (using the TextBlob library in Python). The script analyses a positive or negative sentiment with regard to tweets, resulting in the following analysis:
- Average polarity per hour - Tweets are analysed on polarity (-1 to +1, negative to positive). This results in a line chart that shows average sentiment per hour
- Count of positive and negative sentiment - Tweets are analysed on outspoken negative or positive tweets (<-0,5 or >0,5). This result in a area graph which shows the volume of negative tweets versus positive Tweets
Personally i think this can be a interesting for sentiment prediction for crypto's, let me know what you think.
Results and raw data are available on
www.altcoinanalytics.com.
https://i.imgur.com/PyufPKy.png