Cryptocurrencies predictions using BIG DATA

[cryptobigdata]

Hello everyone,

In the last month we have seen how the crypto market becomes crazy, a lot of people are investing in Bitcoin and others in some altcoins. With the arrival of the holidays, the market is regulating some coin prices but others like ripple are growing to fast.

In this blog, I will try to explain how the cryptocurrencies prices can be observed and "predicted" with big data technologies. For sure, some of you have heard about Google trends and how the chart looks like the real price chart, but there is another source than can predict this prices with more accurate: Twitter big data. In this social network a lot of people are talking or "tweeting" about their investments and price tendencies.

So, we can collect this data and make some analysis to observe the behavior? Let see an example of Ripple [XRP] big data anlysis and the real price chart for day 27 of December (2017):

https://imgur.com/a/4sQQB

Both charts are very similar...now we can think that people interactions in this social network are extremely related with the speculation and market prices.

In the following example, we can see how if the people stop talking about Ripple coin, the coin prices goes down. This chart is for day 28 of December (2017):

https://imgur.com/a/gUAeW

I think that make this anlysis and predictions with user data is a big idea, hope to see the project move forward so this can help us in our investments.

Already donated some BTC coins to the project. In next few days I will publish more information about new conclusions and new charts.

Enjoy the read! contact me for any question!

[ekolet]

You need to run a regression analysis to see if there is a significant correlation. IE how much of the % price change between t and t+1 can be explained by changes in your twitter metric between t-1 and t. Patters mean absolutely nothing.

[cryptobigdata]

You need to run a regression analysis to see if there is a significant correlation. IE how much of the % price change between t and t+1 can be explained by changes in your twitter metric between t-1 and t. Patters mean absolutely nothing.

Of course you need to make your own regression analysis and predictors. I m sure that some analysis mechanisms are coming to work with this data.
But one thing is 100% real, the speculation tendencies make changes in price charts.
https://imgur.com/a/Mq1FG

You are welcome! and very thanks for your comment

[bountyjunkie]

I think big data can work for young coins. The big ones are immune to chatter(mass speculations).

[cryptobigdata]

I think big data can work for young coins. The big ones are immune to chatter(mass speculations).

Yea, if you see the images in the post, both are with Ripple and my previous answer to ekolet was with Bitcoin....I think this 2 are big ones.

[logosobscura]

I think big data can work for young coins. The big ones are immune to chatter(mass speculations).

Not really- most of Bitcoins recent run has been because of chatter- in the mainstream press. Same with its subsequent retrenchment back to closer to $10k. All crypto is highly speculative so highly influenced by chatter- until there is a substantial utility for them, they will remain so.