The Wide World of Scores and Tau

in #tauchain6 years ago (edited)

From Metrics and Sentiment to Scores

Tauchain is a project in development by the sagacious Ohad Asor. Ohad Asor is a mathematician and programming prodigy who designed TML. TML is a meta language which in many ways resembles datalog but with negation as a bonus. TML will use Futamura Projections to produce a Tau compiler written in TML.

Tau is to be a lot like the Cyc project. It will allow for a shared knowledge base which can be reasoned over using deductive style reasoning. It is important to note in my opinion that Tau will also have a social network element based around discussions. It will allow participants to have a “world view” and this “world view” would be the basis behind what could become a “digital representation” of what a person currently thinks. The logical element of Tau would allow for contradiction avoidance and other useful features during discussion.

Scores are the result of metrics. Scores such as loyalty scores or reputation scores can be created by the Tau community agreeing via discussion on the formulas to use. The metrics such as age of the account, the level of engagement such as the various activity measures of the account, these can be tracked. These metrics can be plugged into formulas to create a number of different loyalty scores. In my opinion the market will create these scoring formulas and the best ones will ultimately be used.

The Technical Aspects Behind Scoring

A brief discussion of the technical mechanism behind scoring requires an understanding of the design pattern called the Token Curated Registry and an understanding of the work behind the Enigma Project. The Token Curated Registry allows for a ranked list to be created. These lists can be the result of collaborative filtering (upvotes/downvotes) or these lists can be based on metrics and ranked by scores.

A privacy concern exists where if there is sensitive data being collected to compute these scores then perhaps hackers or other entities might try to access the data. Enigma Project utilizes the Trusted Execution Environment to allow for Secret Contracts to exist. The Trusted Execution Environment is a hardware encrypted computation environment where not even the machine doing the computations can see the data. To put it succinctly, the Secret Contract is an encrypted smart contract. The data being computed is done in such a way that the input are encrypted, the outputs also may be encrypted, and the computations done in the Trusted Execution Environment are inaccessible and undecipherable by the Secret Nodes.

Secret Nodes in Enigma are similar to the witnesses in other networks or the miners in Bitcoin. These Secret Nodes are given computation tasks which they complete. The computation tasks are done in the Trusted Execution Environment which protects the privacy of the data being computed. So for example we could have a Secret Contract which leverages the design pattern of the Token Curated Registry running on a Secret Node, within a Trusted Execution Environment. This would allow for a ranked list of scores from highest to lowest to remain encrypted and private where the input isn’t seen by anyone and the output could merely be a score.

IF YOU LIKE THIS CONTENT, FULL VERSION EXCLUSIVE FOR SUPPORTERS ONLY:
https://www.minds.com/danaedwards/blog/the-wide-world-of-scores-and-tau-930138409499680768

References

Asgaonkar, A., & Krishnamachari, B. (2018). Token Curated Registries-A Game Theoretic Approach. arXiv preprint arXiv:1809.01756.

Asor, O. (2015). About Tau-Chain. arXiv preprint arXiv:1502.04120.

Koriche, F., & Zanuttini, B. (2010). Learning conditional preference networks. Artificial Intelligence, 174(11), 685-703.

Szabo, N. (1997). The god protocols. Satoshi Nakamoto Institute.

Zyskind, G., Nathan, O., & Pentland, A. (2015). Enigma: Decentralized computation platform with guaranteed privacy. arXiv preprint arXiv:1506.03471.

Sort:  

To listen to the audio version of this article click on the play image.

Brought to you by @tts. If you find it useful please consider upvoting this reply.