From "Analogue" Science to AI-powered Digital Science

TL;DR

This paper advocates transitioning to AI-powered digital science through three innovative projects: a blockchain-based research data platform, an AI-friendly publication standard, and automated knowledge extraction from publications.

Contribution

It introduces three novel project ideas to facilitate the shift from traditional to AI-driven digital science, emphasizing their integration and potential impact.

Findings

Proposes a blockchain-based platform for research data attribution.

Designs an AI-friendly standard for publishing research findings.

Develops methods for extracting knowledge from publications using NLP and image recognition.

Abstract

Phase transition from the human-limited, "analogue" way of research enquiry to the silicon-based, artificial intelligence (AI)-powered digital science is forthcoming. To facilitate this transition, I propose three project ideas: 1) CryptoScience platform, aimed to provide tools for endemic digitalization & open access of the research data, with attribution of the work credit of all involved individuals and parties using blockchain technology; 2) Computational Publication Standard, designed for publishing research findings in an AI-friendly format of knowledge graphs; 3) data modelling from publications, meant to extract key pieces of knowledge (entities and their relationships) contained in manuscript publications using natural language processing and image recognition, to enable their incorporation into the digital knowledge databases. These three ideas could be implemented separately by different research groups and institutions, however, they will strongly benefit from each other's synergistic effect.