Kensuke ITO – Page 3 of 4 – Bitcoin & Conceputual Art の愛好家

Ito et al. (2018) “Information Diffusion Enhanced by Multi-Task Peer Prediction”

Abstract: Our study aims to strengthen truthfulness of the two-path mechanism: an information diffusion algorithm to find an influential node in non-cooperative directed acrylic graphs (DAGs). This subject is important because the two-path mechanism ensures only weak truthfulness (i.e., nodes are indifferent between reporting true or false out-edges), which restricts node selection accuracy. To enhance the mechanism, we employed an additional reward layer based on a multi-task peer prediction, where an informative equilibrium provides strictly higher rewards than any other equilibrium in virtually all cases (strong truthfulness). Rewards, which are derived from a comparison of each report, encourage a node to report true out-edges without affecting its own probability of being selected by the original two-path mechanism. We have also experimentally confirmed that our proposed strongly truthful two-path mechanism can sufficiently elicit true out-edges from each node.

【Best Paper Award】at the 20th International Conference on Information Integration and Web-based Applications & Services (iiWAS2018), November 19th to 21st, 2018, Yogyakarta: Grand Mercure Yogyakarta Adisucipto, pp.96-104.

Link to ACM digital library

美術手帖12月号に論考が掲載されました

「P2Pネットワーク上に文脈を形成するプロトコルは設計可能か？」
大学院の外で初めて自分の研究内容について具体的に触れた。
文章の構成や内容は意識的に美術批評的にならないよう心がけた。

Continue reading “美術手帖12月号に論考が掲載されました”

伊東 (2018) 「分散型オラクルの合意形成に対するピア予測法の潜在的有用性」

Abstract: In this paper, we pointed the potential utility of peer prediction method to the existing consensus building in decentralized oracle systems where participants aim to verify the validity of input information to blockchain without relying on a trusted third party (TTP). This is important because, despite the recent expectation of implementing decentralized oracle systems, few discussions have dealt with the incentive design for their consensus building, much less the synergy with peer prediction method. Specifically, we mentioned the followings through the survey of preceding studies: (i) the current predominant method of staking that allows validators to bet the reward tokens has the limitations such as a vulnerability to strategic behavior and a lack of incentive to participate in the verification, (ii) these problems could be solved by peer prediction method which determines the amount of rewards based on the posterior probability distribution on the report of others updated by one’s own report. Peer prediction method can encourage validators to perform proper verification while supplementing the token-based rewards, and thereby can contribute to the realization of the mining mechanism based on subjective review instead of computational resources. On the other hand, several obstacles still remain to propose a practical incentive design, such as the fluctuation of token price that would prevent peer prediction from incentivizing proper verification.

Full article