Data Biasing Removal with Blockchain and Crowd Annotation

Abstract

In today's world of internet marketing product review plays a crucial role. Often E-Commerce companies are accused of manipulating with the data and it suffers from biasing. Biasing leads to wrong customer's view creation thereby prediction for a top rating product fails. This study shows how crowd annotation prevents this biasness and thereby design a smart contract which ultimately will help to eliminate the data manipulation. The smart contract is deployed and tested to see if data can be saved in the run time storage memory regions of the Ethereum blockchain, and the test results are positive. In future this will help to get the unbiased data and better machine learning results, which will ultimately be beneficial for the economy.