科学大数据智能分析软件的现状与趋势

Current Situation and Trend of Intelligent Analysis Software for Scientific Big Data

人工智能领域近年来取得突破进展,如何在自然科学领域采用人工智能新技术促进科学发现,成为科学家和产业界的关注焦点。多学科、跨领域交叉背景下的科学大数据挖掘分析与知识发现,依赖于构建一套高效、易用、可扩展的科学大数据智能分析软件系统,为复杂数据处理、分析、模式提取和知识发现提供学习模型、算法及开发工具支持。文章选取典型科学领域内代表性的智能分析软件系统进行充分的调研,对比分析这类软件的共性和差异,并探讨其发展趋势。在此基础上,文章提出一个面向科学大数据的一体化、可定制的智能分析框架,支撑科学家交互式构建智能分析模型并高效执行,为快速开展科学发现研究提供系统和工具支撑。

The field of artificial intelligence has made a breakthrough in recent years. How to promote scientific discovery in the field of natural science, especially the field of Earth Science with mass and multi-source data, has become the focus of scientists and industry. The scientific data mining analysis and knowledge discovery in the multidisciplinary and cross field intersecting background depend on building a set of efficient, easy to use and extensible scientific data analysis software system for scientific data. It provides learning models, algorithms and development tools for complex data processing, analysis, pattern extraction and knowledge discovery. In this study, the representative intelligent analysis software system in the typical scientific field is selected to make a full investigation and comparison on the generality and difference of this kind of software, and the development trend is also discussed. On this basis, this study proposes an integrated and customizable intelligent analysis framework for scientific big data, which supports the interactive construction of intelligent analysis models, and provides systems and tools supporting for the rapid development of scientific discovery research.