Sentiment analysis for crypto currencies

The bon mot ‘there is no bad press’ does not apply to crypto currencies. EOS had a media presence on Twitter regarding the accounts frozen by block producers, but this was rather negative. Accordingly, it is helpful to follow the general mood on the social media around a crypto currency.

Scionanalytics offers first approaches for such a sentiment analysis

In this, the evaluation of the tweet is classified AI-supported. In a pie chart, the sum of the tweets within the last week is displayed. If you move the mouse over a section of the pie chart, you get a rough picture of the mood within the last week for a crypto currency:

As you can see, the opinion of the Twitter community regarding DASH has been rather positive. Of a total of around 43,000 tweeets, almost 5,000 are positive, while just under 1,400 are negative. The ratio between positive and negative tweets is therefore 3.6.

For EOS, for example, the picture looked somewhat less good last week thanks to the dramas affecting the block producers: Here the ratio is 2.3, so in relative terms fewer positive tweets were posted than at DASH.

An interesting project with optimization potential

In addition to these features, Social Coin Analytics tries to give AI-based price forecasts. Should the probability for such a forecast exceed a certain level, it will be displayed in the dashboard. The Iunera team continues to train the AI in order to make successively better predictions.

Social Coin Analytics from Iunera knows how to please. Data nerds can spend a lot of time studying the tweeting behavior for different crypto currencies.

As with many young projects, working with Social Coin Analytics results in a small wish list in your head. So it would be extremely useful if one could also look at longer time intervals. Furthermore, it would be great if you could also consider data from other social networks. Also important is a filter option in the time-dependent plot so that you can only look at the positive or negative tweets as a function of time. After all, it would be fantastic if you could watch the tweets yourself at certain times, as in the above example on June 24.

Nevertheless: The wishlist shows above all that you quickly get a desire for more when using it. An interesting tool, which one should pursue further.