Fighting fake news with Proof-of-Value
How can we leverage the Blockchain to distinguish and incentivize valuable content?
According to a Stanford University study, over 30,000,000 fake items were shared in the 2016 presidential campaign. Fake news is an extremely complicated and multi-faceted problem, and there is no simple solution. Because of this we have decided to take a multi-pronged approach to the fake news problem.
Proof-of-Value
The Proof-of-Value protocol is our approach to distinguish and reward valuable content while preventing the proliferation of fake news and content that does not align with the values of the community. Distinguishing valuable contributors is a large part of this protocol and in order to accomplish this we will have an extensive reputation system.
Reputation
The reputation system will be a fundamental part of the Proof-of-Value protocol. Users will have a reputation score for each community that they participate in within Sapien. The reputation score is a dynamic quantity that changes as a user’s contributions are evaluated, reflecting their expertise and reliability on a given subject. Users will have to spend significant time and effort contributing to a community to gain reputation within it. Users with a greater reputation will have more influence with a community and will be able to promote valuable content and reduce the spread of content that is opposed to the values of the community.
Machine Learning
Machine learning is a powerful tool in the fight against fake news. Advances in machine learning have pushed the boundaries of natural language processing and enabled machines to discern meaning from text. Discerning Fake news from reality presents an excellent potential application of machine learning, and although there is not yet a simple algorithmic solution to determine truth, several potential foundations to such an algorithm have been proposed. Sapien will initially use a technique called “stance detection” to determine whether claims made in the title of posts and articles agree with the text in the majority of articles. When determining the probability that a claim is fake, sources will be weighted according to their reputation. Our algorithm will determine the probability that a given claim is false and display it to users, allowing them to more readily identify fake news.
Identity Verification
Identity verification is another important issue in reducing the spread of misinformation. Sapien will enable storage of identities on the Blockchain for the purpose of identification. Media publishers and well-known users will also be able to verify their identity on the blockchain, making it significantly more difficult for trolls to deceive users into believing that they are someone they are not.
Something-at-Stake Design
By allowing users to stake SPN tokens, we are able to align the interests of the Sapien platform and users. Users have a vested interest in creating quality content on the platform, or risk losing their staked SPN. With this “something-at-stake” design, trolls are discouraged from malicious behavior. Valuable contributions will fuel the growth of the platform, increase the value of the SPN token, and compensate users for their merits in the form of SPN rewards.
By combining several methods we will create a well rounded protocol to reduce the spread of fake news and promote truly valuable content.
