Search

The high backpressure heating mode of direct air-cooling unit faces limitations due to the return water temperature of heat network and the coordinated heat exchanger unit control of air-cooling island. This limitation complicates the task of achieving a balance between operation economy and power generation output. To solve this inherent contradiction， Ebsilon Professional， a commercial software， is used to calculate and analyze the key parameters of the direct air-cooling exhaust steam extraction heat supply system. These parameters include the backpressure of the air-cooling island， steam turbine output and heat consumption， and suction coefficient of the injector under different loads and variable operating conditions. The calculations reveal that， during low load in winter， the optimal operating backpressure for the heat supply system falls within the range of 13~20 kPa， which allows for the gradient utilization of heating energy， ensures a balance in unit output， and meets the safety requirements for preventing freezing.

" 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.

With the explosive popularity of ChatGPT and the development of AI generated content (AIGC), the new generation of artificial intelligence (AI) technology has triggered people's imagination and discussion about the production mode of digital content and the industry transformation. At the same time, it also prompted more scholars' thinking beyond embracing the technological changes in the field information resource management. Based on this, we invited 6 experts to carry out related discussions from the perspective of the disciplinary construction and development of information resource management. (1) AIGC and the field of library and information service: The essence of the rapid development of AI technology lies in the improvement of people's ability in knowledge acquisition brought by the continuous enrichment of data resources and the improvement of quality. The professionals in the field of library and information service should fully understand their value and their role, give full play to the advantages of data resources in libraries, effectively utilize their expertise in knowledge organization and management, actively promote research and development of AI technology, and contribute their wisdom and solutions. (2) The disruptive change brought about by AIGC: The emergence of ChatGPT has brought subversive challenges to the library and information service field. The professionals in this field should actively promote the transformation of the AI era, seize the opportunity, and explore new growth points around knowledge production, resource discovery, knowledge relevance and scientific research ethics. In addition, they should give play to the advantages of our country's political system and intensive development policies, promote the transformation of information services to knowledge content generation service, and raise the knowledge productivity of the whole society. (3) AIGC and cultural heritage resources: ChatGPT has ushered in an era of model-driven content production and carry out innovation in the mode of intelligent service of cultural heritage resources supported by AIGC. In addition, this paper also analyzes the copyright ownership, real-time, interpretability, verifiability, trustworthiness, algorithm bias and cognitive misdirection of large-scale AIGCs. In the future, universal value rationality such as public order and good customs should be endowed with the AIGC, and the paradigm change in the field of cultural heritage should be actively welcomed and promoted. (4) AIGC and researches on intelligent information processing of ancient books: While ChatGPT plays a positive role in promoting intelligent information processing of ancient books and promoting cultural transmission and civilization inheritance, it also brings challenges in cultural value and national security, content security and information governance, information literacy and humanistic education. It is necessary to make efforts in the aspects like providing higher quality and larger scale of precise processing of ancient books in the future, so as to build a scientific, comprehensive and systematic system of information security and digital ethics. (5) AIGC and intelligent tool revolution: This paper analyzes the three elements of big data, giant model and huge computing power in the development of AIGC technology, as well as the four development directions of large and complete data volume, fine granularity of data analysis, multi-source and multi-mode semantic integration of semantic association, human-computer integration of information services and intelligent interaction. It is pointed out that AIGC will bring about the paradigm change of content production in three aspects: production innovation, automatic generation and ecological cultivation. At the same time, it will also bring challenges in technology monopoly and hegemony, network information security technology integration and other aspects. (6) AIGC and library: Based on the analysis of the core technology innovation trend in the ChatGPT model expansion, data enhancement and model fusion, this paper demonstrates the necessity and crisis of using ChatGPT in libraries. In addition, from five aspects of intelligent management system, retrieval system, recommendation system, question answering system and search engine, this paper puts forward some suggestions on the use of GPT in library service integration and innovation promotion.

[Purpose/significance] This paper conducts a study for the mainstream news media for People's Daily Online corpus, aiming to provide ideas and practical support for the study of automatic text summarization, which can then be applied to news and other related text information processing, and contribute to knowledge aggregation services and information access research. [Method/process] The experimental corpus of this research was the sub-corpus of the People’s Daily Online in January 2015, June 2015 and January 2016 in the new era People’s Daily (NEPD). Based on TF-IDF, Textrank and other extractive automatic summarization algorithms, based on the generative automatic abstractive summarization model for the pointer-generator network, the research was carried out and analyzed and evaluated the summarization results. [Result/conclusion] The experiment builds a news extraction automatic abstractive algorithm the Pointer-Generator Networks model for the People’s Daily corpus, and constructs a network model of news generative automatic summary pointer generation for People's Daily Online corpus. Fruitful experimental results are evaluated by Rouge indicator (including 3 indicators: Rouge-1, Rouge-2 and Rouge-L). This article provides corpus support and practical support for the automatic news summarization system.

[Purpose/Significance] The aim of this study is to reveal the characteristics of our times of important words in the No. 1 Central Document, the first document issued by the Central Committee of the Communist Party of China every year. [Method/Process] Taking the People's Daily corpus processed by manual word segmentation as the research object, the frequency of several important words in the No.1 Central Document of 2020 which have appeared in People's Daily corpus in January since 2015 is calculated. From a diachronic perspective, this paper compares the changes of word frequency in People's Daily in different periods and analyzes the reasons. [Results/Conclusions] First, words about macro top-level decision making policies and common social phenomena are found to have higher frequency in People's Daily corpus. Second, words related to three rural issues, i.e. agriculture, rural area and farmers, do not have higher frequency, especially those related to specific jobs. Finally, it is feasible to use People's Daily corpus as a basis to conduct the diachronic analysis of word frequency of policy making documents.

Objective: To assess the practical and health economical values of non-invasive prenatal test (NIPT) in Changsha Municipal Public Welfare Program., Methods: A retrospective analysis was carried out on 149 165 women undergoing NIPT test from April 9, 2018 to December 31, 2019. For pregnant women with high risks, invasive prenatal diagnosis and follow-up of pregnancy outcome were conducted. The cost-benefit of NIPT for Down syndrome was analyzed., Results: NIPT was carried out for 149 165 pregnant women and succeeded in 148 749 cases (99.72%), for which outcome were available in 148 538 (99.86%). 90% of pregnant women from the region accepted the screening with NIPT. 415 (0.27%) were diagnosed as high risk. Among these, 381 (91.81%) accepted amniocentesis, which led to the diagnosis of 212 cases of trisomy 21 (PPV=85.14%), 41 cases with trisomy 18 (PPV=48.81%) and 10 cases with trisomy 13 (PPV=20.83%). The sensitivity and specificity of NIPT for trisomy 21, trisomy 18 and trisomy 13 were (97.70%, 99.98%), (97.62%, 9.97%) and (100%, 99.97%), respectively. In addition, 213 and 30 cases were diagnosed with sex chromosomal aneuploidies (PPV=46.2%) and other autosomal anomalies (PPV=16.57%), respectively. For Down syndrome screening, the cost and benefit of the project was 120.79 million yuan and 1,056.95 million yuan, respectively. The cost-benefit ratio was 1: 8.75, and safety index was 0.0035., Conclusion: NIPT is a highly accurate screening test for trisomy 21, which was followed by trisomy 18 and sex chromosomal aneuploidies, while it was less accurate for other autosomal aneuploidies. The application of NIPT screening has a high health economical value.