Crypto asset taxonomy classification and crypto news sentiment analysis

Date

2020-9

Author

Köse, Ozan

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

57
views

171
downloads

Cite This

There are over 1900 cryptocurrencies trading in cryptocurrency exchanges as of June 2020 and the number is rapidly growing. In the current crypto scene, cryptocurrencies are seen as investment vehicles by many, yet every crypto asset is designed to operate in a specific sector within a pre-defined business model. There are many characteristics that differentiate one crypto asset from another and there are numerous internal and external factors that affect each crypto asset differently depending on these characteristics. Crypto investors can leverage these factors and characteristics and use these indicators to create different trading strategies. In this thesis, to guide the investors, we classify the crypto assets under various characteristics and provide sentiment analysis on crypto related news. As our first major task, we focus on automated annotation of the cryptocurrencies in terms of sector, transaction anonymity and asset type through the public information. To this aim, we generated an annotated dataset by collecting information from various sources. The collected dataset includes cryptocurrency descriptions annotated with sector, asset type and transaction anonymity labels. For this task, we utilized a divide-and-conquer supervised learning approach and compared its performance against several supervised learning algorithms for the three aspects. As our second major task, we focus on automated sentiment analysis of crypto asset related news on news title, summary and content. For this, we have generated an annotated news dataset that is collected from various public news sources. We use this dataset to fine-tune a successful neural network model that is trained on financial sentiment analysis task and compared its performance to various dictionary-based methods.

Subject Keywords

Cryptocurrency Document Classification, Crypto News Sentiment Analysis, Sector Classification, Asset Type Classification, Transaction Anonymity Classification

URI

https://hdl.handle.net/11511/69285

Collections

Graduate School of Natural and Applied Sciences, Thesis