Crypto Whitepaper Similarity
We finally have our similarity model working! We tested this on a large set of whitepapers, and later manually clustering them to measure accuracy of the model. We are fairly happy with the current state, however, would know better once we run it on large set.
Similarity algorithm works by first finding key terms in a whitepaper, and later checking with all the other whitepapers to find similar key terms. If you feel a bit technical, here’s a good start: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.332.4480&rep=rep1&type=pdf
As the first post(a lot of interesting ones to follow :)), we analyze cryptos which have fairly large user group in reddit, telegram, and other sources. Next, we’ll run it on upcoming ICOs, and all also lesser known alts which are not that well analyzed on social media forums.
Please do let us know which coins/tokens you want us to run the similarity measures on next. You can post on our subreddit post https://www.reddit.com/r/shamcoin/comments/7sxo5z/finding_similarity_in_cryptos/
Lighter the color, more similar it is. From this set, we couldn’t find any tokens significantly similar to each other(BLT, ENG, STORM are close). Also, District0x(DNT) is the most unique one 🙂