Computational Scientific Discovery: Reflections on Its Aims and Methodology

doi:10.15936/j.cnki.1008-3758.2024.02.002

Abstract

Abstract: Computational scientific discovery arises from the convergence of artificial intelligence and science. It computationally reconfigures the cognitive processes of human scientific discovery and realizes the technological enhancement of natural human cognitive abilities， enabling appropriately programmed computers to simulate human discovery activities and assist the mankind in acquiring new knowledge of the objective world. Computational scientific discovery introduces computation as a method and perspective into the studies on scientific discovery， revealing finer structures in the past modes of reasoning and providing new insights into classic problems. Through methodological reflection， it becomes evident that the limitations of computational science discovery depend on the extent to which one can characterize specific scientific discovery activities as computationally tractable problems， design efficient algorithms for these problems， provide ample and accurate scientific data， and establish reasonable and necessary trust in the discovery system.

Key words: computational scientific discovery; artificial intelligence; scientific discovery; computation; methodology

CLC Number:

N031

PU Jianghuai. Computational Scientific Discovery: Reflections on Its Aims and Methodology[J]. Journal of Northeastern University(Social Science), 2024, 26(2): 10-17.

References

[1]Newell A，Simon H A. GPS， a program that simulates human thought［M］∥Feigenbaum E A，Feldman J. Computers and Thought. New York: McGraw-Hill， 1963:279-293.
[2]Simon H A. Machine discovery［J］. Foundations of Science， 1995，1(2):171-200.
[3]Deroski S，Langley P，Todorovski L. Computational discovery of scientific knowledge［C］∥Deroski S，Todorovski L. Computational Discovery of Scientific Knowledge: Introduction， Techniques， and Applications in Environmental and Life Sciences. Berlin: Springer， 2007:1-14.
[4]Wajnberg C D，Corruble V，Ganascia J G，et al. A structuralist approach towards computational scientific discovery［C］∥Suzuki E， Arikawa S. Discovery Science. Berlin: Springer， 2004:412-419.
[5]Sozou P D，Lane P C R，Addis M，et al. Computational scientific discovery［M］∥Magnani L， Bertolotti T. Springer Handbook of Model-based Science. Cham: Springer， 2017:719-734.
[6]孟小峰. 科学数据智能: 人工智能在科学发现中的机遇与挑战［J］. 中国科学基金， 2021，35(3):419-425.
[7]Jumper J，Evans R，Pritzel A，et al. Highly accurate protein structure prediction with AlphaFold［J］. Nature， 2021，596:583-589.
[8]Schulte-Sasse R，Budach S，Hnisz D，et al. Integration of multiomics data with graph convolutional networks to identify new cancer genes and their associated molecular mechanisms［J］. Nature Machine Intelligence， 2021，3:513-526.
[9]Bartlett L K，Pirrone A，Javed N，et al. Computational scientific discovery in psychology［J］. Perspectives on Psychological Science， 2022:17456916221091833.
[10]Giza P. Automated discovery systems， part 1: historical origins， main research programs， and methodological foundations［J］. Philosophy Compass， 2022，17(1):e12800.
[11]Simon H A. Scientific discovery as problem solving［J］. International Studies in the Philosophy of Science， 1992，6(1):3-14.
[12]Langley P，Simon H A，Bradshaw G L，et al. Scientific discovery: computational explorations of the creative processes［M］. Cambridge: MIT Press， 1987:8.
[13]Lenat D B. The ubiquity of discovery［J］. Artificial Intelligence， 1977，9(3):257-285.
[14]Langley P. Data-driven discovery of physical laws［J］. Cognitive Science， 1981，5:31-54.
[15]Holland J H，Holyoak K J，Nisbett R E，et al. Induction: processes of inference， learning， and discovery［M］. Cambridge: MIT Press， 1989:326.
[16]Brannigan A. Artificial intelligence and the attributional model of scientific discovery［J］. Social Studies of Science， 1989，19:601-613.
[17]Langley P. The computational support of scientific discovery［J］. International Journal of Human-Computer Studies， 2000，53(3):393-410.
[18]郭贵春，郝宁湘. 丘奇-图灵论点与人类认知能力和极限［J］. 齐鲁学刊， 2004(5):65-70.
[19]Buchanan B G，Feigenbaum E A. DENDRAL and meta-DENDRAL: their applications dimension［J］. Artificial Intelligence， 1978，11(1/2):5-24.
[20]Valdés-Pérez R E. Algorithm to infer the structures of molecular formulas within a reaction pathway［J］. Journal of Computational Chemistry， 1994，15(11):1266-1277.
[21]刘晓力. 计算主义质疑［J］. 哲学研究， 2003(4):88-94.
[22]董春雨，薛永红. 机器认识论何以可能?［J］. 自然辩证法研究， 2019，35(8):3-10.
[23]Korb K B. Introduction: machine learning as philosophy of science［J］. Minds and Machines， 2004，14(4):433-440.
[24]杨仁杰. 因果发现与归纳问题［J］. 自然辩证法通讯， 2020，42(8):18-25.
[25]张怡. 论数据归纳推理［J］. 哲学分析， 2021，12(6):125-136.
[26]Giza P. Automated discovery systems and scientific realism［J］. Minds and Machines， 2002，12(1):105-117.
[27]保罗·汉弗莱斯. 延长的万物之尺——计算科学、经验主义与科学方法［M］. 苏湛，译. 北京: 人民出版社， 2017:149.
[28]尼古拉斯·雷舍尔. 复杂性——一种哲学概观［M］. 吴彤，译. 上海: 上海科技教育出版社， 2007:180-190.
[29]闫坤如. 可解释人工智能: 本源、进路与实践［J］. 探索与争鸣， 2022(8):102-109.
[30]Charlie Wood. Powerful “machine scientists” distill the laws of physics from raw data［EB/OL］. (2022-05-10)［2022-09-20］. https:∥www.quantamagazine.org/machine-scientists-distill-the-laws-of-physics-from-raw-data-20220510/.