Application Status and Development Trend of Artificial Intelligence in Sci-tech Periodical Publishing
浏览次数: 817次 Release time: 2019/10/24 9 : 16 : 05Views : 817
share to:

Application Status and Development Trend of Artificial Intelligence in Sci-tech Periodicals Publication Abstract: AI technology has a wide application in the publishing industry. It combines artificial intelligence technology and sci-tech periodicals publishing, and formulates corresponding expert systems, intelligent assistant editors, processing and decision support systems, etc. The comprehensive social and economic benefits of the sci-tech journal publishing industry will be improved. This article summarizes the research progress and application status of artificial intelligence technology in various fields of sci-tech journal publishing at home and abroad, and elaborates the development trend of artificial intelligence technology application and research in the fields of publishing process management and peer review. Innovation and reform provide a reference.
Keywords: artificial intelligence; scientific and technical journals; publication; natural language processing; artificial intelligence (AI) is an emerging discipline that has achieved remarkable development in many fields, and many countries have improved artificial intelligence To the height of national strategy. China emphasizes the promotion of the development of artificial intelligence and reaching the world's leading level. 2016-2017 released the "Internet + Artificial Intelligence Three-year Action Implementation Plan" [1], "New Generation Artificial Intelligence Development Plan" [2], "Three-year Action Plan to Promote the Development of New Generation of Artificial Intelligence (2018-2020)" [3], and in the "Thirteenth Five-Year Plan" National Science and Technology Innovation Plan [4], "13th Five-Year Plan" National Strategic Emerging Industry Development Plan [5], "19th National Congress Report" [6] and other governments The report proposes encouraging the development of the artificial intelligence industry. The "New Generation Artificial Intelligence Development Plan" [2] states that "Big data-driven knowledge learning, cross-media collaborative processing, human-machine collaborative enhanced intelligence, group integrated intelligence and autonomous intelligent systems have become the focus of artificial intelligence development. The new stage. "The United States emphasizes the maximization of artificial intelligence innovation and utility, and reduces negative impacts [7-9]. The focus includes investing and developing artificial intelligence that is beneficial to society, and understanding and solving the ethical, legal, and sociological impact of artificial intelligence. , Through standards and benchmarks to measure and evaluate artificial intelligence technology, to better understand the human needs of artificial intelligence in the country, etc. The United Kingdom emphasizes the protection of personal data security [10-11]. The key content includes the combination of effective machine learning and the threat to basic personal freedom rights such as personal data security and privacy. Establish clear rules to ensure the artificial intelligence machine decision-making process. Transparency and accountability, maximizing the socio-economic benefits of technological progress and minimizing its potential threats.
On December 2, 2017, China's first and world's first open road driverless bus officially launched. The Hong Kong-Zhuhai-Macao Bridge officially opened on October 23, 2018. Face recognition using artificial intelligence technology for immigration clearance can achieve customs clearance as fast as 30 seconds. With a voice ticket vending machine in the Shanghai subway station, passengers only need to say their destination, and the ticket vending machine will automatically recommend routes and stops. In the 2017 Double Eleven Shopping Festival, about 170 million shopping posters were automatically designed by Ali's "Luban" artificial intelligence design system. The "unmanned bank" that has appeared in the past two years has no tellers, security guards, and lobby managers. Most of the conventional personal business is replaced by intelligent robots. Complex businesses can provide remote teller consulting services through video systems. Artificial intelligence is sweeping every aspect of our lives in an overwhelming trend, and the press and publishing industry is no exception. Artificial intelligence technology has been applied to content creation [12-14], evaluation [15], and communication [16]. In the field of academic publishing, artificial intelligence technologies [17] that complete complex tasks such as plagiarism review, statistical analysis and inspection, etc. have also begun to emerge. This paper summarizes the research progress and application status of domestic and foreign artificial intelligence technology in the field of sci-tech journal publishing in the field of content discovery, production, processing, and review. The paper describes the application and research of artificial intelligence technology in the areas of publishing process management and peer review The development trend aims to provide basic theoretical references for technological innovation and institutional reform of sci-tech journals.

1. The basis and conditions of applying artificial intelligence technology in the publishing field of scientific and technological journals The role of artificial intelligence technology cannot be separated from three major factors: algorithms, computing power, and data. The core of realizing artificial intelligence is algorithm, that is, computer program, just like the recipes that guide humans to cook dishes. The recipes write the ingredients, order of ingredients, and fire control to make a dish. The algorithm tells the computer to do What, how. In addition, the computer that executes the algorithm needs certain computing capabilities, especially for computers that execute artificial intelligence algorithms. In recent years, the increasing use of artificial intelligence is also related to the dramatic improvement in computer performance. In addition, human intelligence is related to data. Humans are intelligent because they can receive data, process data, and feed back data. In the era of the mobile Internet, the amount of data generated and accumulated by humans is an unimaginable magnitude. Now we are discharging data all the time, and collecting and recording data is more convenient than in any previous era. In the 1990s, shortly after the birth of the World Wide Web, academic publishing giants such as Elsevier, Springer, and Wiley established online electronic databases of journals, and realized structured processing of article content very early. The data format of an article conforms to a format that can be read by a machine, coupled with the increasing number of papers or data every year. With such a data foundation, the application of artificial intelligence technology in the field of sci-tech journals becomes logical.
In recent years, various translation software have been more accurate and logical for the translation of entire sentences. This is due to the breakthrough in natural language processing technology [18], a class of algorithms for processing human languages in artificial intelligence technology, which is also an important condition for the application of artificial intelligence in the field of sci-tech periodicals.

2 The application of artificial intelligence technology in the publication of scientific and technological journals, the continuous development of computing power, and the continuous accumulation of data mean that artificial intelligence technology can play an increasingly greater role in our work. In the field of publishing scientific and technological journals, artificial intelligence can not only accelerate academic exchanges, but also overcome human prejudices. Artificial intelligence technologies applied to related issues such as academic content search, plagiarism detection, and identification data falsification are subverting the field of scientific journal publishing.
2.1 Content Discovery According to statistics, about 3 million English papers are published every year worldwide, that is, one article every 10 seconds [19]. Researchers and journal editors can hardly keep up with this speed, even in their own narrow subject areas. Therefore, many tools that use artificial intelligence technology to efficiently search, classify, review, construct knowledge maps, and drill deep into massive academic literature have appeared. ScienceOpen [20], Semantic Scholar [21], Text & data mining, Yewno [22] , Mendeley Suggents, Sparrho, Meta, Nano, Iris are examples of such tools. The biggest feature of this type of academic search engine is that they all use artificial intelligence technologies such as intelligent information capture, semantic technology, personalized search, and content understanding based on deep learning, which is more conducive to helping users screen useful information and improve academic Information search and filtering efficiency.
2.2 Content Production In the news industry, artificial intelligence technology has long been used in content production for reporting facts and data [12-14]. In the field of scientific journal publishing, many content production tools have also appeared. The most famous one is the "SCIgen" developed by researchers at the Massachusetts Institute of Technology in 2005. Its purpose is purely entertaining. Later, someone used SCIgen to generate a paper, with Kim Kardashian and Satoshi Nakamoto as authors. , Published in the academic journal of Drug Designing & Intellectual Properties International Journal [23]. Although it is a spoof, it is enough to prove that a computer program is completely possible to produce a standard academic paper. In addition, PaperRobot, the latest thesis generation software developed by researchers at Rensselaer Polytechnic Institute and the University of Washington, generates papers in the field of biomedicine. After Turing test, the abstracts, conclusions and future work are only 30%. In%, 24%, and 12% of cases, human experts believe that PaperRobot generates better writing than humans [24]. With the advancement of technology, more serious tools for academic content production have appeared in the field of sci-tech journal publications, with representative examples being SCInote [25] and articoolo [26]. Among them, SCInote can link user-defined content to form a draft of a formal academic paper based on the data in the reference and the experimental process. Articoolo only needs a user to give a topic that can be described in 2-5 words, and can automatically "write" an article. Although artificial intelligence has been able to write papers, machines still don't have the ability to analyze results and make opinions. As scientific research focuses more and more on data, more dull and templated papers will appear.
2.3 Content Processing Content processing includes the use of artificial intelligence technology to assist humans in reading and judging content, and converting academic content into machine-readable formats to achieve more applications of artificial intelligence technology. UNSILO [27] started in medical journals. The main services provided include: using AI technology such as natural language processing and machine learning to analyze manuscripts, automatically selecting key concepts and summarizing the content of the manuscript, thereby assisting editors or reviewers to judge the manuscript. . UNSILO provides a set of intelligent text extraction and analysis system to understand, manage, classify and index the content of articles, and use artificial intelligence to establish the relationship between content, readers and knowledge points. After continuous improvements and upgrades, UNSILO has integrated many new features: analyzing the paper, proposing which keywords best represent the argument or research findings, and indicating whether these claims are similar to previously published papers, allowing editors to understand the research results faster .
2.4 Content Review Content review is a form of artificial intelligence applied to review manuscripts. Statreviewer and Statcheck are two representatives of AI technology applied to content review of papers. The functions of the two applications are similar. The result is correct. Springer Nature has applied Statreviewer in the review system of some of its journals. However, there has been no literature report on the use experience and effect analysis of these two applications.

3 Trends in the application of artificial intelligence technology in scientific and technological journals Algorithms and computing power will continue to increase, and data will continue to accumulate. The creation of metadata, automatic translation, machine-generated content, machine-assisted quality inspection and other AI technologies are critical to the publishing process of future scientific and technological journals. However, the publishing industry of sci-tech periodicals emphasizes the rigor of science and accurate documentation. Therefore, it is the direction of future development of AI technology in the field of sci-tech periodicals publishing that can truly understand the technology required by researchers and scientists. In 2017, an article in the well-known American popular science magazine Wired [27] discussed whether AI technology could replace the currently controversial peer review method in academic publishing. If AI can really develop to the level of writing a standard academic paper, can artificial intelligence also review manuscripts? Inspired by a writing robot, several researchers at the University of Trieste have written a tool that automatically generates review reports. The tool evaluates any paper positively, negatively, or neutrally based on a pattern set by the user. When scientists mixed machine-generated reviews with real-life reviews for a group of human readers, a quarter of computer-generated reviews were considered more useful than human ones. In addition, there have been similar successful applications in the legal field, such as LawGeex's SaaS (Software as a Service) platform, which uses AI technology to overcome legal obstacles in the review and approval process before daily commercial contract signing. Therefore, in the future, AI technology will also be applied to peer review of academic papers, but just as lawyers will not lose their job, the decision-making power of peer review is still in the hands of human reviewers. AI brings to improve or accelerate the process, and grasp The mistakes that humans must devote a lot of effort to catch, the trained AI can achieve good performance in specialized and specific fields, and the cooperation between AI and professionals is the irreversible trend in the future.
In addition to the emergence of transformative technologies that change the publishing model of sci-tech periodicals, AI-based journal products, services, and market segments in the field of sci-tech journal publishing will also emerge. For example, editing processing mainly relies on the editing experience of editors and understanding of specifications, and deals with the language and format specifications of manuscripts. This process has operational characteristics such as formal normalization and logical verifiability, which is also good at AI. In the commercial environment, there are already very mature AI robots that deal with similar process tasks. Therefore, the emergence of automatic proofreading software or applications based on AI technology to aid editing and processing will improve the quality of article editing and reduce the pressure on editing.

4 Conclusion Artificial intelligence is bringing great opportunities to the transformation and upgrading of sci-tech journal publishing. At the same time, artificial intelligence will also have a huge impact on the publishing of sci-tech journals, but it does not exaggerate and rely on artificial intelligence to human intelligence. In order to achieve a win-win situation between artificial intelligence and human development, intelligent innovation in the field of sci-tech journal publishing, and sustainable development can be achieved by coordinating the two and complementing and integrating applications.
[1] Three-year implementation plan of "Internet +" artificial intelligence [J]. China Information Technology Education, 2016 (11): 2-3.
[2] Development plan of new generation artificial intelligence [J]. Science & Technology Review, 2018, 36 (17): 113.
[3] "Three-year Action Plan to Promote the Development of the New Generation of Artificial Intelligence Industry (2018-2020)" [J]. Robot Industry, 2018 (1): 70-79.
[4] "Thirteenth Five-Year Plan" National Science and Technology Innovation Plan [J]. Science and Technology Innovation and Productivity, 2016 (9): 2, 121.
[5] "13th Five-Year Plan" National Strategic Emerging Industry Development Plan (full text) [J]. China Strategic Emerging Industries, 2017 (01): 57-81.
[6] Xi Jinping. Definitely building a well-off society in an all-round way and capturing the great victory of socialism with Chinese characteristics in the new era [N]. People's Daily, 2017-10-28 (001).
[7] White House report: Prepare for the future of artificial intelligence [J]. Information Security and Confidentiality of Communications, 2016 (12): 9-14.
[8] He Zhe. Towards the era of artificial intelligence: Also on the direction of American artificial intelligence strategy and its reference to Chinese artificial intelligence strategy [J]. E-Government Affairs, 2016 (12): 2-10.
[9] Zhang Qian. US White House: Artificial Intelligence, Automation and Economic Reports [J]. Hangzhou Science and Technology, 2017 (2): 57-60.
[10] Li Ruirui. Library Service Model Reconstruction and Innovation Development from the Perspective of Artificial Intelligence: An Analysis Based on the Report of British "Artificial Intelligence: Opportunities and Impacts of Future Decisions" [J]. Books and Information, 2017 (6): 30 -36.
[11] Wang Xiaodi. British Intelligent Robot Technology Innovation Research and Development Trend Analysis [C] // Beijing Institute of Science and Technology Information. 2018 Beijing Science and Technology Information Society Academic Annual Meeting: Intelligent Technology Development Intelligence Service Advances "Proceedings of the Forum. Beijing Science Institute of Technical Information: Beijing Institute of Scientific and Technical Information, 2018: 11.
[12] Han Qin. Comparative research on the development of news robots at home and abroad [J]. Publishing reference, 2019 (6): 33-36.
[13] Liu Xin, Shao Lu. Research on the Development and Application of Robot News [J]. Western Broadcasting and Television, 2017 (16): 34-35.
[14] Shi Anbin, Long Yifan. Traceability, Status Quo and Prospect of News Robots [J]. Young Journalists, 2016 (22): 77-79.
[15] Lu Qian, Ren Yuanyuan. Subversion or assistance? : The Practice and Reflection of "News + Artificial Intelligence" [J] .Young Journalists, 2018 (30): 75-77.
[16] Wang Chenxi. Talking about the influence of artificial intelligence writing robots on news communication [J]. Research on Communication Power, 2018, 2 (33): 33, 35.
[17] Zhang Yong, Wang Chunyan, Wang Xiying. Artificial Intelligence and the Future of Academic Journal Editing and Publishing [J]. Chinese Editor, 2019 (4): 64-68.
[18] Fang Mingzhi. Development and Future of Natural Language Processing Technology [J]. Science & Technology Communication, 2019, 11 (6): 143-144.


Disclaimer: This website is a non-profit website. Most of the information comes from the recommendations of netizens. If the author or other copyright owner believes your rights are violated, please let us know and we will remove them within 24 hours.