The development of contemporary science and technology puts forward extremely high requirements for data and computing power. In the field of traditional basic technology research, the research teams in research institutes and universities generally lack the support ability in this respect. In order to cope with this change, the "open science platform" has gradually become popular in Europe and America in recent ten years, providing scientific research communities with various supports including open data and related analysis tools, open computing platforms and open source code. In the era of competition among big countries in science and technology, "open science platform" has gradually become a key variable to promote the development of science and technology. China needs to increase investment in this field in order to realize the follow-up and operation of basic science platform construction in the stage of scientific paradigm transformation.

1

The law behind scientific progress: from paradigm to fourth paradigm

What is the track of scientific progress?

In the early days, it was generally believed that scientific development was gradual, and scientists emerged constantly, which promoted the development of existing scientific theories and pushed scientific discoveries closer to the truth.

1962, thomas kuhn, an American scholar, put forward in his book The Structure of Scientific Revolution that, taking the historical development of physics as an example, there is a regular discontinuity in scientific development. The book says that science consists of a series of alternating "normal scientific stages" and "scientific revolution" stages. In the stage of scientific revolution, with the breakthrough of conceptual level, the existing theoretical framework has been completely subverted and reconstructed by the new scientific paradigm, such as the transformation from Newtonian mechanics to quantum mechanics.

Kuhn created a new concept of philosophy of science "paradigm shift". In the paradigm shift theory defined by Kuhn, scientific discovery follows another five-stage circular path:

1. "Pre-paradigm stage": Several incompatible and incomplete theories gradually converge into a dominant paradigm generally accepted by the scientific community.

2. "Normal scientific stage": under the background of the dominant paradigm, we constantly solve scientific problems and promote scientific progress.

3. "Crisis stage": The abnormal phenomena that can't be solved by the dominant paradigm will accumulate for a long time, which will eventually lead to the trust crisis of the dominant paradigm.

4. "Paradigm transformation stage": After the major failure of the dominant paradigm, the basic assumptions of scientific research are re-examined and a new paradigm is established.

5. "New Paradigm Stage": After the scientific revolution, the dominant position of the new paradigm was established.

Kuhn started a new discipline, called the history of science or sociology of science. Sociologists and historians began to regard science as a common subculture. It is generally believed that paradigm shift has promoted scientific development.

After Kuhn, Jim Gray, an academic leader in the computer field and a legendary scientist in the database field, summarized the development history of science and technology into four paradigm stages:

1. first normal form: empirical evidence. In the history of civilization for thousands of years, the vast majority of human scientific and technological development stems from the observation and experimental summary of natural phenomena. Compared with the unquestionable dogma in religion or theology, the empirical paradigm based on observation and experiment is a big step in the development of human science and technology.

2. The second paradigm: theoretical science. Compared with first normal form's dependence on observation and experiment, the scientific theory of the second paradigm needs to "know why" and explain some natural laws in principle. Therefore, the theory must meet further requirements, such as falsifiability's predictive ability and the consistency of multiple independent sources.

3. The third paradigm: computational science. With the increasing computing power of computer, it is gradually applied to scientific research. Different from the first two paradigms, a brand-new technical tool-computational model and system simulation has emerged in computational science. Based on massively parallel computer architecture, this tool uses the computing power of computer to simulate complex processes, such as atmospheric circulation, nuclear reaction process and virus infection process, by designing algorithms and programming. It is widely used in fields such as economics, psychology and cognitive science that lack simple and intuitive analytical solutions.

4. The fourth paradigm: data exploration. From gene sequencer, deep space telescope to social media covering the whole world, human society has been deeply digitized. Every day, in the digital earth where we live, a huge amount of data is generated every day. It is estimated that by 2025, the world will generate 463 EB of data every day, equivalent to 212,765,957 DVDs, reaching as high as 30 Everest. In such an era of data explosion, the technological revolution based on data is also taking place, and data science has become the frontier of technological development. A series of scientific research methods, such as data-centered computing platform, data processing, processing and sharing tools, algorithms and model bases, constitute the fourth paradigm of technological development.

2

"Open Science Platform": Better Meeting the Challenge of the Fourth Paradigm

In the era of science and technology dominated by the fourth paradigm, the research team should not only study the cutting-edge research papers, but also need the experimental data sets involved in these papers in order to reproduce the experimental results. Huge data and huge computing demand bring new challenges to innovation. Traditional scientific communication is limited to papers and limited public data sets. In order to promote the sustainability of technological development in the new paradigm era and accelerate the cooperation and sharing of scientific research teams, Europe and the United States began the "open science" movement:

ArXiv: Paul Ginsberg, Ph.D. in Physics, 199 1 A shared arxiv knowledge base was established in Los Alamos National Laboratory, USA, to help global scientific research teams share the full-text digital version of preprinted research papers. In 2008, arXiv broke through 500,000 milestones; By the end of 20 14, it had reached1000000; Today, arXiv receives more than 15000 contributions from the global scientific research community every month. Some publishers also grant arXiv authors the right to archive peer-reviewed printed copies for free. ArXiv also encouraged more scientific research institutions, such as the National Library of Medicine, to establish a PubMed central knowledge base in the medical field after ArXiv.

Papers with codes: Compared with arXiv, papers with codes go further in the field of open access. This platform not only shares papers, but also encourages scientific research teams to share codes and data sets corresponding to paper work. At the same time, the platform provides a related evaluation environment, which can rank the submitted algorithm codes on public data sets. The paper with code not only creates an open collection of papers, codes and data sets, but also provides a free and open technology evaluation method platform, which significantly promotes the technology sharing and cooperation of scientific research teams in related fields and accelerates the process of scientific and technological innovation.

NASA's Open Source Science Program: In order to lower the threshold of space science exploration and let more scientific research teams participate in NASA's scientific research, NASA proposed the Open Source Science Program (OSSI). The goal of OSSI is to establish an open scientific community within ten years. The participating scientific research teams promise to open up and * * enjoy relevant software, data and knowledge (including algorithms, papers, documents and related auxiliary information) during the research and development process, and make full use of technical tools such as collaboration tools and cloud computing to improve the speed and quality of NASA's scientific progress.

three

Science and technology game among big countries: the importance of opening up scientific platform

In the era of the fourth paradigm, the explosive growth of data plays an extremely important role in scientific research and development, and also brings new challenges to scientific research teams in various disciplines, especially in the non-computer field, involving the acquisition and sharing of data and models, the scale expansion of computing models and system simulations, the construction of large-scale synthetic data, tools for multimodal data processing and analysis, data extraction, conversion, loading, data visualization and interaction, artificial intelligence tool sets and computing platforms.

Today, the new round of competition among big countries is largely reflected in the competition between the ecosystem supporting the development of science and technology and the innovation system, and the open science platform has become one of the important basic capabilities to promote scientific research and development. The new generation innovation system driven by the fourth paradigm has already had the first-Mover advantage in Europe and America, and it has had ten years of experience from government-led scientific research projects to academic practice. China needs to realize the deficiency in this field, speed up the construction of an independent and open scientific platform, and realize the follow-up and operation of scientific and technological development in the transformation stage of scientific paradigm.

References:

1./books/2065438+02/07/03/Structure-Science-Revolution-Kuhn-Comment

2./ Biography/Thomas S-Kuhn

3./en-us/research/publication/ fourth paradigm-data-intensive-scientific discovery/

5. https://en.wikipedia.org/wiki/metaphysics _ (Aristotle)

6. https://www.weforum.org/agenda/2019/04/How much data is generated every day -cf4bddf29f/

7 . https://www . skatelescope . org/8 https://arxiv . org/help/stats/202 1 _ by _ area

end