切实发挥科技文献在AI4S中的知识底座作用
TP18%G350%G250; "人工智能驱动的科学研究"(AI for Science,AI4S)是一场正在发生的科技革命,是将人工智能技术与科学研究深度结合,帮助促进发现新知识、解决科学难题的新型科学研究范式.随着AI4S的研究推进,其发展脉络、机遇和挑战、需求和任务、实现路径等问题值得进一步讨论.为此本刊邀请了7位专家组织了本期笔谈.1)支撑AI4S的知识服务:AI4S对当下的知识服务提出了更高的要求,包括多层次知识发现与获取需求;跨学科研究和创新需求;以用户为中心的参与式服务需求,使得知识服务场景向多元化,智能化,专业化,个性化转变.为此须重新定位AI4S环境中知识...
Saved in:
| Published in | 农业图书情报学报 Vol. 35; no. 10; pp. 9 - 13 |
|---|---|
| Main Author | |
| Format | Journal Article |
| Language | Chinese |
| Published |
中国农业科学院农业信息研究所
05.10.2023
中国科学院文献情报中心,北京 100190 国家新闻出版署学术期刊新型出版与知识服务重点实验室,北京 100190 中国科学院大学经济与管理学院信息资源管理系,北京 100190 |
| Subjects | |
| Online Access | Get full text |
| ISSN | 1002-1248 |
| DOI | 10.13998/j.cnki.issn1002-1248.23-0850 |
Cover
| Abstract | TP18%G350%G250; "人工智能驱动的科学研究"(AI for Science,AI4S)是一场正在发生的科技革命,是将人工智能技术与科学研究深度结合,帮助促进发现新知识、解决科学难题的新型科学研究范式.随着AI4S的研究推进,其发展脉络、机遇和挑战、需求和任务、实现路径等问题值得进一步讨论.为此本刊邀请了7位专家组织了本期笔谈.1)支撑AI4S的知识服务:AI4S对当下的知识服务提出了更高的要求,包括多层次知识发现与获取需求;跨学科研究和创新需求;以用户为中心的参与式服务需求,使得知识服务场景向多元化,智能化,专业化,个性化转变.为此须重新定位AI4S环境中知识服务的新角色,明确其在全面支撑科技创新进程中的新任务,树立大文献观,兼顾普惠与专深,以支撑跨学科创新.2)建设AI4S的知识底座:人工智能的本质是知识的获取与利用,而科技文献则是人类知识的主要载体.中国科学院文献情报中心充分认识到AI带来的科研范式变革,提出了构建AI4S的科技文献知识底座的概念,积极挖掘科技文献中蕴含的科学知识和高质量数据,努力构建AI4S的领域智能知识底座,将"科技文献库"转变为"科技知识引擎",支撑AI4S所需的查询循证、态势感知、推理预测、生成启示等智能化服务.3)驱动AI4S的科学数据:科学数据有效聚合为发挥AI4S的强大功能奠定了数据基础,是图书馆实现AI时代角色与功能变革的前提,是推动科研服务转型、深化科研支持、加速科技创新的必要条件.目前图书馆有效聚合科学数据为AI4S提供支撑仍面临宏观和中观上的诸多挑战,应对该挑战有以下实现路径:明确图书馆在科学数据管理中的角色与作用;营造科学数据管理环境;构建科学数据管理合作网络;提升科学数据管理服务能力.4)AI4S与古典文献智能语言模型:AI4S技术能够用于文献和文本的分析,更快速、更全面地理解大量的历史文献和文化资料.古典文献智能语言模型是人工智能技术在古籍文献研究领域的一项重要突破,为古典文献研究带来了新的机遇和挑战.随着多模态、生成式GPT模型的流行,AI4S情境下古典文献智能语言模型将更加注重整合多样信息、提高适应性、增强知识表示和服务于更广泛的应用场景.5)面向AI4S的图书馆数字学术服务:基于LLM的AI4S和AIGC推动智慧图书馆建设的理念不谋而合,给图书馆数字学术服务带来了机遇和挑战.基于AI4S平台化趋势与数字学术服务中台化特征适配,以及图书馆界长期服务 |
|---|---|
| AbstractList | ChatGPT和GPT4的出现，让人们深刻认识到人工智能时代已经来临。AI所表现出的学习知识、掌握知识、应用知识解决问题的能力让世人为之瞠目。在科学界，各个学科领域都在积极利用人工智能助力于科学研究，构建智能驱动的科研模式（AI for Science，AI4S）。由于科技文献蕴含和保存了人类知识，在人工智能时代来临之时，科学界普遍认识到了科技文献对AI4S的重要性。 TP18%G350%G250; "人工智能驱动的科学研究"(AI for Science,AI4S)是一场正在发生的科技革命,是将人工智能技术与科学研究深度结合,帮助促进发现新知识、解决科学难题的新型科学研究范式.随着AI4S的研究推进,其发展脉络、机遇和挑战、需求和任务、实现路径等问题值得进一步讨论.为此本刊邀请了7位专家组织了本期笔谈.1)支撑AI4S的知识服务:AI4S对当下的知识服务提出了更高的要求,包括多层次知识发现与获取需求;跨学科研究和创新需求;以用户为中心的参与式服务需求,使得知识服务场景向多元化,智能化,专业化,个性化转变.为此须重新定位AI4S环境中知识服务的新角色,明确其在全面支撑科技创新进程中的新任务,树立大文献观,兼顾普惠与专深,以支撑跨学科创新.2)建设AI4S的知识底座:人工智能的本质是知识的获取与利用,而科技文献则是人类知识的主要载体.中国科学院文献情报中心充分认识到AI带来的科研范式变革,提出了构建AI4S的科技文献知识底座的概念,积极挖掘科技文献中蕴含的科学知识和高质量数据,努力构建AI4S的领域智能知识底座,将"科技文献库"转变为"科技知识引擎",支撑AI4S所需的查询循证、态势感知、推理预测、生成启示等智能化服务.3)驱动AI4S的科学数据:科学数据有效聚合为发挥AI4S的强大功能奠定了数据基础,是图书馆实现AI时代角色与功能变革的前提,是推动科研服务转型、深化科研支持、加速科技创新的必要条件.目前图书馆有效聚合科学数据为AI4S提供支撑仍面临宏观和中观上的诸多挑战,应对该挑战有以下实现路径:明确图书馆在科学数据管理中的角色与作用;营造科学数据管理环境;构建科学数据管理合作网络;提升科学数据管理服务能力.4)AI4S与古典文献智能语言模型:AI4S技术能够用于文献和文本的分析,更快速、更全面地理解大量的历史文献和文化资料.古典文献智能语言模型是人工智能技术在古籍文献研究领域的一项重要突破,为古典文献研究带来了新的机遇和挑战.随着多模态、生成式GPT模型的流行,AI4S情境下古典文献智能语言模型将更加注重整合多样信息、提高适应性、增强知识表示和服务于更广泛的应用场景.5)面向AI4S的图书馆数字学术服务:基于LLM的AI4S和AIGC推动智慧图书馆建设的理念不谋而合,给图书馆数字学术服务带来了机遇和挑战.基于AI4S平台化趋势与数字学术服务中台化特征适配,以及图书馆界长期服务 |
| Abstract_FL | "AI for Science"(AI4S)is a new scientific research paradigm that deeply integrates AI technology with scientific research to promote the discovery of new knowledge and the solution of scientific problems.As the application of AI4S in the natural sciences and humanities and social sciences advances,its development line,opportunities and challenges,needs and tasks,and ways of realization deserve further discussion.In order to advance AI4S research,promote scientific and technological(S&T)innovation and progress,and facilitate the effective strengthening of the discipline of information resources management,our journal has invited seven experts to organize this academic conversation on AI4S.1)Supporting knowledge services for AI4S:In the current landscape of intelligent knowledge services,the requirements for supporting AI4S have increased,including the need for multi-level knowledge discovery and acquisition,cross-disciplinary research and innovation,and user-friendly participatory services.In addition,knowledge service scenarios are moving towards diversification,complexity,depth,specialization,and personalization in ubiquitous knowledge discovery,generative content services,and multi-round interactive service exploration.In response,professional science and technology information organizations need to reassess the role of knowledge services in the AI4Science environment and their significance in comprehensively supporting the S&T innovation process.This involves establishing a broad literature perspective,deepening full-text knowledge elements,balancing universal and specialized depth,autonomously developing core products,and deeply engaging with professional fields to support interdisciplinary innovation.2)Building the knowledge base of AI4S:The essence of artificial intelligence(AI)lies in the acquisition and use of knowledge,and scientific and technological(S&T)literature is the primary carrier of human knowledge.Fully recognizing the paradigm shift in scientific research brought about by AI,the Document Information Center of the China Academy of Sciences has proposed the concept of building a S&T literature knowledge base for AI4S.It is actively exploring the scientific knowledge and high-quality data contained in the S&T literature,strives to build a domain intelligent knowledge base for AI4S,and transforms the"S&T literature database"into a"scientific knowledge engine"that supports intelligent services such as query evidence,situational awareness,inference prediction,and generation of insights required by AI4S.3)Powering AI4S with scientific data:Effective aggregation of scientific data is the foundation for unleashing the powerful capabilities of AI4S.This is essential for libraries to adapt their roles and functions in the AI era and is a crucial prerequisite for catalyzing the transformation of scientific research services,deepening scientific research support,and accelerating S&T innovation.Currently,libraries face various macro and meso challenges in effectively aggregating valuable scientific data to provide support for AI4S.To address these challenges,the following ways can be pursued:defining the roles and functions of libraries in scientific data management;promoting a conducive environment for scientific data management;establishing a collaborative network for scientific data management;and enhancing the service capacity of scientific data management.4)AI4S and intelligent language modeling for classical literature:AI4S technology can be used to analyze documents and texts,enabling a faster and more comprehensive understanding of a vast amount of historical documents and cultural materials.The development of intelligent language modeling for classical literature represents a significant breakthrough in the field of ancient literature research,bringing new opportunities and challenges.With the increasing popularity of multimodal and generative GPT models in the context of AI4S,the intelligent language modeling of classical literature will focus on integrating diverse information,enhancing adaptability,improving knowledge representation,and addressing a wider range of application scenarios.5)Library Digital Scholarly Services for AI4S:The concept of using LLM-based AI4S and AIGC to drive the development of smart libraries is consistent with the vision for digital scholarly services in libraries,and presents both opportunities and challenges.Given the trends towards AI4S platformization and the characteristics of"middle-end"digital scholarly service,as well as the longstanding tradition of libraries in serving scholarly research,the reengineering path for the library's digital scholarly services platform includes three approaches:building an AI4S service platform independently,purchasing and utilizing third-party AI4S platforms,and promoting embedded knowledge services as a component of scientific intelligence.This innovative approach addresses the dilemmas of financial resources,human resources,cognitive and practical gaps,and emphasizes the importance of user needs in the AI4S environment.It also focuses on knowledge organization and service delivery to meet user needs in the AI4S landscape.6)Historical evolution and logical structure of the scientific intelligence paradigm(AI4S):AI4S is a scientific paradigm change dominated by the full application of AI technology to various disciplines,and its logical structure includes"data+model"-driven,knowledge ecology created by machine conjecture,and application scenarios expanded by algorithmic thinking.In the era of digital civilization,AI4S-driven scientific progress and social development must carry forward the value of science and technology for the good,effectively select the theoretical arguments and proposals for extending AI4S to the field of social sciences and humanities,and improve the series of mechanisms for integrating human decision-making and machine intelligence.7)Development opportunities and prospects of AI4S in the era of generative AI:With the advances in generative AI,pre-training algorithms and large-scale pre-trained models have provided significant opportunities for AI4S in various disciplinary domains.These technologies have shown immense potential and value for applications in diverse fields such as industrial inspection,robotics,and medicine.Additionally,it is crucial to emphasize the importance of key factors such as the constraints of technical implementation conditions for large pre-trained models,the sustainability of data/computing resources,and the transparency,fairness,and accessibility of the technology. |
| Author | 张智雄 |
| AuthorAffiliation | 中国科学院文献情报中心,北京 100190;中国科学院大学经济与管理学院信息资源管理系,北京 100190;国家新闻出版署学术期刊新型出版与知识服务重点实验室,北京 100190 |
| AuthorAffiliation_xml | – name: 中国科学院文献情报中心,北京 100190;中国科学院大学经济与管理学院信息资源管理系,北京 100190;国家新闻出版署学术期刊新型出版与知识服务重点实验室,北京 100190 |
| Author_xml | – sequence: 1 fullname: 张智雄 organization: 中国科学院文献情报中心,北京100190 |
| BookMark | eNo9j89LAkEcxedgkJV_RXTosNt3ZnZ3Zo8q_RCkCD3USeaHG2qM1BTVrUNkhBBCdQipk-glu4bQf-O6_RltGJ0evPfhPd4Sypi2qSO0hsHFNAz5RtNVptVwG9YaDEAcTDzuEuoA9yGDsv_eIspZ25DgY44h8L0sYvFdJx6_xg-9WXeQDHuz--vZcyfpjuP-KF_yKtPP9-TlJnkbfH_cxpOneDKcfvWTx9EKWojEsa3n_nQZVbc2q8Udp7y3XSrmy44K0n3lKwhFGBFSh1D7fsBDrKSg3NOEBhEJCXDJhFbC0xqDJFxqXzFJGNMCGKPLaH1eeyFMJMxRrdk-PzXpYM1cndkTedkiQGj6D2jKrs5ZJWyaW6tru4fVyn7hoPALwZz6AbFAahA |
| ClassificationCodes | TP18%G350%G250 |
| ContentType | Journal Article |
| Copyright | Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| Copyright_xml | – notice: Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| DBID | NSCOK 2B. 4A8 92I 93N PSX TCJ |
| DOI | 10.13998/j.cnki.issn1002-1248.23-0850 |
| DatabaseName | 国家哲学社会科学文献中心 (National Center for Philosophy and Social Sciences Documentation) Wanfang Data Journals - Hong Kong WANFANG Data Centre Wanfang Data Journals 万方数据期刊 - 香港版 China Online Journals (COJ) China Online Journals (COJ) |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| EndPage | 13 |
| ExternalDocumentID | nytsqbxk202310003 NYTSQBXB2023010003 |
| GrantInformation_xml | – fundername: 国家重点研发计划 funderid: (2022YFF0711900) |
| GroupedDBID | -0Y 2B. 2C- 7X2 92D 92I ABUWG ACGFS AEUYN AFKRA ALMA_UNASSIGNED_HOLDINGS ALSLI ATCPS BENPR BHPHI CCEZO CCPQU CNYFK CUBFJ CW9 DWQXO GROUPED_DOAJ HCIFZ M0K M1O NSCOK PIMPY TCJ TGH 4A8 93N ABJNI PHGZM PHGZT PMFND PSX |
| ID | FETCH-LOGICAL-c623-c5c09a9f22e09d556891cba384d236f29208b7adca4dd10b28bd5c7b277da0773 |
| ISSN | 1002-1248 |
| IngestDate | Thu May 29 03:54:13 EDT 2025 Tue Jan 21 20:55:46 EST 2025 |
| IsDoiOpenAccess | true |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 10 |
| Keywords | 科学数据聚合 knowledge base intelligent language model for classical literature scientific and technological literature 科技文献 古典文献智能语言模型 intelligent knowledge service scientific data aggregation 知识底座 图书馆数字学术服务 AI4S 智能知识服务 library digital scholarly services |
| Language | Chinese |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c623-c5c09a9f22e09d556891cba384d236f29208b7adca4dd10b28bd5c7b277da0773 |
| OpenAccessLink | http://dx.doi.org/10.13998/j.cnki.issn1002-1248.23-0850 |
| PageCount | 5 |
| ParticipantIDs | wanfang_journals_nytsqbxk202310003 cass_nssd_NYTSQBXB2023010003 |
| PublicationCentury | 2000 |
| PublicationDate | 2023-10-05 |
| PublicationDateYYYYMMDD | 2023-10-05 |
| PublicationDate_xml | – month: 10 year: 2023 text: 2023-10-05 day: 05 |
| PublicationDecade | 2020 |
| PublicationTitle | 农业图书情报学报 |
| PublicationTitle_FL | Journal of Library and Information Science in Agriculture |
| PublicationYear | 2023 |
| Publisher | 中国农业科学院农业信息研究所 中国科学院文献情报中心,北京 100190 国家新闻出版署学术期刊新型出版与知识服务重点实验室,北京 100190 中国科学院大学经济与管理学院信息资源管理系,北京 100190 |
| Publisher_xml | – name: 中国农业科学院农业信息研究所 – name: 中国科学院大学经济与管理学院信息资源管理系,北京 100190 – name: 中国科学院文献情报中心,北京 100190 – name: 国家新闻出版署学术期刊新型出版与知识服务重点实验室,北京 100190 |
| SSID | ssib051810654 ssib001102652 ssib023646129 ssib046786149 ssj0001811708 ssib051370036 ssib036275894 ssib042363015 |
| Score | 2.3308406 |
| Snippet | ChatGPT和GPT4的出现，让人们深刻认识到人工智能时代已经来临。AI所表现出的学习知识、掌握知识、应用知识解决问题的能力让世人为之瞠目。在科学界，各个学科领域都在积极利用人工智能助力于科学研究，构建智能驱动的科研模式（AI for... TP18%G350%G250; "人工智能驱动的科学研究"(AI for... |
| SourceID | wanfang cass |
| SourceType | Aggregation Database |
| StartPage | 9 |
| Title | 切实发挥科技文献在AI4S中的知识底座作用 |
| URI | https://www.ncpssd.cn/Literature/articleinfo?id=NYTSQBXB2023010003&type=journalArticle https://d.wanfangdata.com.cn/periodical/nytsqbxk202310003 |
| Volume | 35 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwnV1LaxRBEG7yAPEiiorRGEKwTzJxXt3TfZzZnSUKBiQrrKcwPbNjJDCiuwHNyYMYkYAE1IMEPYXkYrxKwH-T3fVnWNXz2MEEiV6G3n5UdVX145vZrmpCbomuTBUMHIMlCTdcV8VGJExlWEryNO3y1ErQOfn-Ml966N7rsM7E5ELt1NJGXy3Gm6f6lfyPVSEP7Ipesv9g2YooZEAa7AtPsDA8z2RjGjIqBIX3eUj4IZWhzmlRadGQU9GgPqOhR32vzPHxZAMkJNetPF1Ht5KQEP5dd4WGLg0E9ZtYLKGBqxMtTQvyW1RwbBD4VLIiAQywVROpYGUXaNVhr-4W16WaOJBFlgENQp0jqc91_xwqWNFRXxOHbhRFmFMOD80WOpzLIrELoUR6-UVw5WcMOz8Qx6qBV5MtZ9-sqadiJpGkb5-mp5NdrBEMWlhqN7S0DSpzpUC1xm2MQZVfWFrsAPltL3n4z3KLyCOqlFPBrC34soYccqfaE3sSQEChN6U4W3-imVQ8FkEPGDNwvBlXRySzl_3eM_ViHXWF_8E4k2Ta9jw8ijDtB82gNQa9gBE5q0Aq3gjAayDawTjUYrzqAn7mTg0Ewk4pAKNV9ZnleBilaPxbWGYZdE9_uUTfZDP3MS3kOEdoKemdv8kJuCeGd0LtIJelUfa4huXaF8mF4iVs3s9n1CUysbl2mXiDt1uDwy-D9zvD7b3R_s7w3avhp63R9uFg9wDnxfGPb6PPr0df9359fzM4-jg42j_-uTv6cHCFtFthu7FkFNeKGDFgfSNmsSkjmdp215QJBuCTVqwiR7gJaCbF29uE8qIkjtwksUxlC5Ww2FOg_CQyPc-5Sqayp1n3Gpl3pHBgE1RW17TcVDKhUtfjTCoADdKO1AyZRXFXs14vWV1-1F55EHQCNKipLTpDFgo1rBaLClT90-zXz1LpBjk_nlOzZKr_fKN7E8ByX80Vo2VOf2z6DZxFjrk |
| linkProvider | EBSCOhost |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=%E5%88%87%E5%AE%9E%E5%8F%91%E6%8C%A5%E7%A7%91%E6%8A%80%E6%96%87%E7%8C%AE%E5%9C%A8AI4S%E4%B8%AD%E7%9A%84%E7%9F%A5%E8%AF%86%E5%BA%95%E5%BA%A7%E4%BD%9C%E7%94%A8&rft.jtitle=%E5%86%9C%E4%B8%9A%E5%9B%BE%E4%B9%A6%E6%83%85%E6%8A%A5%E5%AD%A6%E6%8A%A5&rft.au=%E5%BC%A0%E6%99%BA%E9%9B%84&rft.date=2023-10-05&rft.pub=%E4%B8%AD%E5%9B%BD%E7%A7%91%E5%AD%A6%E9%99%A2%E6%96%87%E7%8C%AE%E6%83%85%E6%8A%A5%E4%B8%AD%E5%BF%83%2C%E5%8C%97%E4%BA%AC+100190&rft.issn=1002-1248&rft.volume=35&rft.issue=10&rft.spage=9&rft.epage=13&rft_id=info:doi/10.13998%2Fj.cnki.issn1002-1248.23-0850&rft.externalDocID=nytsqbxk202310003 |
| thumbnail_s | http://utb.summon.serialssolutions.com/2.0.0/image/custom?url=http%3A%2F%2Fwww.wanfangdata.com.cn%2Fimages%2FPeriodicalImages%2Fnytsqbxk%2Fnytsqbxk.jpg |