Scrap news articles/blogs related to ‘Cryptocurrencies’ from the internet and apply Text-Mining approach to :
1. Identify and Visualize Co-occurrences/ Relationships across different named entities such as Location, Organization and People from the large text corpus.
2. Automatically gauge the Sentiment across different Use-cases/verticals of Cryptocurrencies
2. Objective:
1. Identify and Visualize Co-occurrences/ Relationships across different
named entities such as Location, Organization and People from the
large text corpus.
2. Automatically gauge the Sentiment across different Use-
cases/verticals of Cryptocurrencies.
3. Analyze general Cryptocurrency Trend across Region, time and
different Cryptocurrencies.
4. Interpretation:
1. By clicking on a particular named Entity, all the
co-occuring People, location and Organizations
can be Analysed using Domain knowledge.
2. For example, in the graph shown on left , Co-
occurrence network of Ripple Co-founder 'Brad
Garlinhouse' is shown.
3. Note how closely he is related with the famous
Vitalik Buterin , the founder of Ethereum.
4.Since Ripple is based in California, hence the
evident relationship in the graph.
5. Also, Ripple has recently confirmed the
expansion of its business with China's Central
bank 'People Bank of China'. The same is evient
from the graph shown.
6. Many more useful relationships can be
extracted from the Co-occurrence graph.
6. Cryptocurrency Trend by Region
Cryptocurrency Popularity trend is high in many European countries like
Switzerland, Slovenia, Greece and Germany. Also trending in US, UK , Malaysia, India
and Singapore.
8. Financial Sentiment Over Past 1.5 years
There have been regular ups and downs in the general cryptocurrency sentiment. The largest jump was around
December 2017 end to Mid-January 2018. This is evident both from the Sentiment across News articles and and
Google Keywords trend Analysis for keywords such as ‘Bitcoin’, ‘BTC’ , ‘ETH’, ‘LTC’ .
9. Findings
• ‘Banking’, ‘Payments’ and ‘Startups’ - are the top three Cryptocurrency
verticals.
• There have been regular ups and downs in the general cryptocurrency
sentiment. The largest jump was around December 2017 end to Mid-
January 2018. This is evident both from the Sentiment across News articles
and and Google Keywords trend Analysis for keywords such as ‘Bitcoin’,
‘BTC’ , ‘ETH’, ‘LTC’ .
• Cryptocurrency Popularity trend is high in many European countries like
Switzerland, Slovenia, Greece and Germany. Also trending in US, UK ,
Malaysia and Singapore.
• Big players in Crypto space ,identified through NER Co-occurrence Analysis
– IBM, Microsoft, JPMorgan and many other.