Ancient books, that is, the books and documents produced and published by engraving and copying before 1912, are used to inherit civilization, popularize education and record history, carrying heavy history and culture. Relevant statistics show that of the more than 2.7 million ancient books we have completed the census, only more than 70000 can be read online. More massive cultural resources of ancient books need to be digitized, which is also one of the important topics for the protection, inheritance and opening of ancient books. At present, the "Huidian · ancient book digital service platform" launched by Shanghai ancient book publishing house has attracted the attention of the industry. The platform aggregates the high-quality ancient book resources of publishing houses in Shanghai, the Yangtze River Delta and even wider regions, and uses the latest optical character recognition (OCR), natural language processing, large-scale corpus, machine learning punctuation and other ancient book intelligent algorithm technologies, Build a knowledge service platform for traditional culture and ancient books industry. "In the north, there is the 'Jihe network' of Zhonghua publishing house, and in the south, there is the 'Huidian' of Shanghai ancient books publishing house." Yang Guanghui, deputy curator of Fudan University Library and executive vice president of the Chinese Academy of ancient books protection, said that the digitization of ancient books plays a positive role in promoting the inheritance, protection and utilization of ancient Chinese civilization. On the one hand, this platform can popularize the ancient books and cultural resources accumulated by Shanghai ancient books publishing house for many years to the public through digitization, On the other hand, it can also accelerate the process of digital transformation and publication of ancient books and promote the digital development of the corresponding publishing industry. The digitization of ancient books has a long way to go after decades of development The 14th five year plan of Shanghai for comprehensively promoting urban digital transformation points out that it is necessary to "deepen the construction of cultural big data system and promote the digitization of cultural resources", and realize the inheritance and innovation of excellent traditional culture with digital technology under the background of media integration and knowledge integration. Digitization of ancient books is the direction of protecting and making rational use of ancient books. In an interview with reporters, Shi Xiang, a researcher at the Institute of Ancient Books Collation of Fudan University, said, "there is a contradiction between the 'use' and 'possession' of ancient books. Everyone wants to turn them over. Over time, it will inevitably affect the protection of ancient books". After digitization, the "mother book" of ancient books will not have to "throw its head and show its face" at various risks. At the same time, the ancient books "raised in the boudoir but not known" can go out of the "boudoir" after digitization, which can meet the reading needs of more readers without time and geographical restrictions, and realize the change of one to many, point to point and virtual to reality. From the "acquisition side" of "turning paper books into electronic scanning edition", to the "production side" of "turning electronic scanning edition into text edition", and then to the "application side" of "turning text edition into ancient book research and learning system", the digitization process of ancient books is not complex. In decades of historical development, there are two watersheds. First, in the 1980s, Chinese American scholar Chen bingzao proposed using computers to count the words of a dream of Red Mansions, and computer technology and humanistic research gradually began to combine. The second is that in 1999, the electronic version of "Si Ku Quan Shu" by Wen yuange, known as the "model book of large-scale Chinese electronic publishing project", came out. The digital library of ancient books in China has achieved more than 30000 achievements in the development of the "Digital Library of ancient books"; The "Chinese classic ancient books library" of Zhonghua publishing house has released more than 3000 point school-based ancient books with 1.5 billion words; The "Chinese basic ancient books library" of irusheng company receives 10000 kinds of books, including both full texts for retrieval and original images of ancient books. But at the same time, the road of promoting the digitization of ancient books is also covered with thorns. On the one hand, the reason lies in the cost of ancient books. According to Hou Junming, head of the digital publishing center of Shanghai Ancient Books Publishing House, "most ancient books are expensive to obtain. In addition, the costs of production, copyright, platform development and copyright protection technology research and development are high. Relatively speaking, the return cycle of ancient books digitization is relatively long." On the other hand, according to the requirements of the national general survey of ancient books, the identification and cataloging of all ancient books, including book title, volume number, author, version, volume storage, volume number, collection and printing, must be clarified one by one. The workload is huge, and the professional level of cataloging appraisers is very high. In fact, there are still a large number of existing digital resources of ancient books in China, which are black-and-white images with low resolution, which are difficult to meet the needs of readers and researchers. Accelerate the digital transformation of ancient books and integrate the massive knowledge system of ancient books with new technologies Lu Jian, chief editor of Shanghai Ancient Books Publishing House, said that ancient books sorting is an ancient cause, while digitization belongs to the present, and digital transformation represents the forefront of the industry. At a time when ancient books are in urgent need of digitization, the emergence of "Huidian · ancient books digital service platform" is like a dawn. The development of OCR system, automatic punctuation and automatic indexing technology of this platform has achieved initial results. OCR technology can quickly identify a book with an accuracy of 93%. After the machine punctuation with ideal accuracy is achieved, the remaining difficult problems can be completed quickly through experts and scholars, freeing scholars from a large number of simple and repetitive work. With the support of OCR text generation technology, natural language processing text sorting and indexing and other advanced technologies, a large number of excellent ancient books can face the public and professional researchers on an accurate and authoritative platform, greatly improve the use efficiency, and effectively promote the convenient, fast and effective use of a large number of excellent traditional cultural information. This platform also analyzes the massive ancient book resource content of Shanghai ancient book publishing house, reveals its knowledge structure, reconstructs the original organization form of ancient book content, and creates a new knowledge module to realize the knowledge-based and professional service of ancient book resources. Through the blessing of technology, the text in ancient books can be quickly "live" from the depth and breadth of content available. Hou Junming said: "Using new technology to integrate the knowledge system of a large number of ancient books and deeply explain the historical origin, development context and basic trend of Chinese culture will help to promote the construction of ideological system, academic system and discourse system with Chinese heritage and Chinese characteristics. From the perspective of regenerative protection, the digitization technology of ancient books is of great significance to the popularization, research and inheritance of Chinese excellent traditional culture Righteousness. " Promoting the digitization process of ancient books is not only the digitization of content, but also the digitization of thinking. Editing, printing and publishing are the basic processes of traditional publishing. In the process of digital project practice, these traditional workflow are also gradually receiving the positive influence from digital thinking. In Yang Guanghui's eyes, the iterative development of ordinary scanning technology to three-dimensional high-definition scanning, the transformation of the Internet to the Internet of things, the digitization of ancient books to Digital Humanities evolution, the books hidden in the library can open the gap between virtual and reality through digital publishing technology, and the formation of "meta universe" of traditional collection resources through new media is not out of reach. (outlook new era)