Empowering Culture and Education Through Digital Content Creation, Preservation, and Dissemination

Abstract

Digital content can bring many advantages to a sustainable world, such as higher accessibility, flexibility, reduction in natural resource consumption, or storage issues. Probably one of the biggest advantages is related to the possibility of preserving any type of document without the use of typography machines, as digital content does not suffer from deterioration. This approach can prove to be extremely useful in the case of important documents for human culture and history. Our paper presents some of the most successful cases of digitization of physical papers, including books, newspapers, and old manuscripts. The analysis of international models of digital libraries is followed by a case study of digitization in Romania, with an analysis of their future perspectives on sustainability.

1. Introduction

Living in a highly technologized world started to be reflected in many domains, even those with hundreds of years of tradition, such as medicine, education, or literature. Evolution can be seen in the dissemination of digital content, in terms of e-books, newsletters, manuals, comics, slides, and many more. Culture and education can be preserved if important documents for human history are archived and protected from natural destruction over time. Digitization thus becomes essential, as a process of creating digital content from physical books, newspapers, or old manuscripts and storing it electronically [1]. Digitization comes with many advantages, including the possibility of preserving any type of document and preventing deterioration. It also assures higher accessibility [2] and flexibility [3], and it can solve ecological problems, as paper is no longer used for creating physical copies. Rare documents become available to a large number of people from all around the world using an Internet connection. Digitization thus becomes a tool for sustainability, reducing paper consumption (and natural resources), deforestation, energy and water consumption, as well as the waste and carbon footprint caused by the typography process. The environmental impact of printed books is enormous, with over 32 million trees cut down annually in the U.S. and over 40 million metric tons of CO₂ emitted globally on book printing [4]. Digitization not only leads to a considerable reduction in these numbers but can also encourage remote learning and ensure knowledge transmission to future generations. Although the production of an e-reader consumes approximately 168 kg of CO₂ emissions, they are used by avid readers, and it is calculated that reading more than 22 books on an e-reader makes the device more sustainable than having the same 22 books printed [5]. In addition, less passionate readers usually use other multifunctional devices, such as smartphones, tablets, or laptops.

The digitization process stands as the pillar of a circular economy model, where waste can be minimized and resources used in an intelligent manner, as opposed to traditional models where resources are taken, industrially processed to be used, and finally disposed. The aforementioned reduction in paper, ink, and energy lead to a decrease in deforestation, while the forwarding of resources to the production of digital books and e-readers allows the continuity of product usage, leading to their extended life cycle. Digital books will not wear out and can be updated permanently, reducing waste, and can be easily accessed by multiple users. Platform-based distribution aligns with the principles of circular economy, encouraging sharing the products/accessing them instead of owning non-reusable physical copies.

By using document image analysis systems, entire collections of important documents can be translated into digital content, as these systems can extract information accurately, almost as if they were being processed by humans. Many types of documents can benefit from the advantages of digitization, including book pages, postal cards, sheet music, geographic maps, or technical drawings. The digitization process, however, is a thorough task, as document image analysis systems are not the only ones involved in the digitization of collections. Old manuscripts, with special characters and many decorations, are difficult to be digitized automatically [6] and often require the intervention of humans. Thus, digitization is often a collaboration between authors, publishers, librarians, researchers, and local governments.

In this paper, we first present the history of the digitization of library resources, including the evolution of digital content and the techniques and algorithms used for digitization. Then, we analyze international models of digital libraries, including global-level collaborations, such as Google Books and Europeana. Finally, we show the Romanian case study, in terms of normative aspects and past and present digitization projects.

2. History of the Digitization of Library Resources

2.1. Evolution—From Microfilms to Digital Libraries

Contemporary large-scale digitization projects, such as Europeana.eu, or Google Books, create an illusion of transparency regarding digital libraries. For the average user, they may be identified as a classical archive enhanced by technological means. However, regardless of some robust advancements, there are still a lot of pending issues. First of all, there is a fierce dispute concerning the proper name to be used; should we speak of a digital or electronic library, or is the notion of a hybrid library or library without walls more appropriate? It seems that, for the moment, it is impossible to find a single concept. “Every library is different, every digital library is different, and different players are advancing many definitions for the digital library” [7]. The second aspect concerns the history of the development of digital libraries, a history that is not a linear tale, starting from some embryonic distant origin but rather an ambiguous and heterogeneous path. In fact, one may consider that the real history of digital libraries is the history of some information systems, very heterogeneous in scope and functionality.

Pragmatically speaking, the digitization process is seen from two complementary angles: storage space and document accessibility. Those two elements were also the main focus in a short paper of Paul Otlet, one of the founding figures of documentation science and Robert Goldschmidt, called “Sur une forme nouvelle du livre: le livre microphotographique” [8]. The main idea of this essay is not only that existing books and documents should be converted into microfilms in order to increase accessibility, but also that the whole industry of printing should switch to microfilm technology. In the early 1920s, Paul Otlet tried to test his ideas by attempting to create an encyclopedia printed completely on microfilm, named the “Encyclopaedia Microphotica Mundaneum”. In 1934, Paul Otlet was tempted by another quasi-utopian project, a so-called Radiated Library, where people would place a telephone call requesting any kind of document from a library. In response, librarians would send the pages as electrical signals to the users [9]. Those revolutionary ideas were not meant to remain only a utopian dream. In 1945, Vannevar Bush, who at that moment was the head of the Office of Scientific Research and Development, devised “a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility” [10]. The big transition took place in 1965, when Joseph Carl Robnett Licklider anticipated, in his book, “Libraries of the Future” [11], that computers would soon be sufficiently powerful to support an electronic library network of numerous remote users.

Nevertheless, despite the rapid advancement of information technology, digital libraries did not really begin to grow until the 1980s. In 1984, The National Archives and Records Administration in Washington initiated one of the first digitization programs. The project, known as the “Optical Digital Image Storage System” (ODISS), aimed to digitize about 220,000 pages of administrative records, using digital methods for reproducing, processing, and retrieving archival records.

Another example of an early digitization project was Spain’s “Archivo General de Indias”, a 200-year-old document archive depicting the history of the Spanish empire in the Americas. Between 1986 and 1998, 11 million pages were digitized, preventing the degradation of original documents in the study rooms.

In 1991, a new digitization program appeared that switched attention to a more specialized public audience. Initially referred to as the “e-print database” and now known globally as arXiv [12], it was created as an experimental way of making scientific communication more productive, collecting online research papers from various fields such as Physics, Mathematics, Computer science, Biology, Economics, and Statistics. In addition to arXiv, it is worth noting initiatives, such as the Electronic Thesis and Dissertations repositories launched in 1996 or the archives of Cognitive Sciences papers, in 1997, which facilitated the development of the “Networked Digital Library of Theses and Dissertations”, an international organization which has the goal of promoting electronic preservation and dissemination of theses and dissertations [13].

The Minerva Project (“Ministerial NEtwoRk for Valorizing Digitization Activities” [14]), which also provides guidance on best practices in digitization, has presented an annual review report on the digitization efforts of all member states of the European Union (EU), for the period 2002–2007.

DELOS, a “Network of Excellence on Digital Libraries”, was founded to provide a forum where researchers and practitioners can exchange ideas and experiences to promote international collaboration in the field of digital library science.

The Bricks Project, which aims to integrate existing resources into a shared digital library that serves as a collective European memory, is another important initiative. Bricks’ goal is to integrate a broad group of users, art practitioners, researchers, as well as teachers, residents, visitors, etc., to share knowledge and services on digital content.

2.2. Technical Aspects—The Evolution of Algorithms, Techniques, and Interoperability

Digitization of content generally follows a workflow consisting of selection, capturing, conversion, preservation, and accessing [15,16]. Selection means choosing targets of future conversion based on their relevance, their condition, and the demand from patrons. Capturing is usually performed via scanning, i.e., translating a target to a basic digital image file; it may be considered part of the conversion. Conversion is more than scanning; conversion means adjusting the image file, extracting text, and adding metadata. In the conversion phase, text is usually extracted using OCR (Optical Character Recognition) techniques. Preservation is concerned with storing the digital content with concern to its integrity and authenticity, structuring it into collections (thus enabling easy access and querying for information), and using interoperable formats. Access is the availability of new digital content to users and patrons. Access control policies dictate which content is available to which users. In the new digital age, access generally requires either:

• the presence of a web interface for downloading or directly viewing content; or
• the integration of specialized services providing the access interface and devices and applications for the end-user (such as Libby [17] from OverDrive).

While there is content “born digital”, i.e., created directly in digital form (no analog format), the focus of the current paper is on analog content that is digitized to its format and its storage (preservation) and user accessibility for searching and filtering.

Recommendations for digitization generally include using a large resolution for the master version (300 ppi, 600 ppi, 800 ppi, and 1200 ppi, depending on scanned content, 24-bit color) and a lossless image format, generally TIFF (Tagged Image File Format) [18,19,20]. Metadata is added to the digitized content to make it useful; descriptive metadata such as title, author, and notes are used for the discovery of a digital object [21]. A metadata schema is generally used as a common way to store fields describing the digital asset; Dublin Core is among the most used schema for metadata storage [22]. Another standard is METS (Metadata Encoding and Transmission Standard), the result of the METAe EU’s project [23].

Recent focus has been given to the digitization of 3D content, its scanning with proper parameters, and proper storage facilities [24]. The evolution of digitization, including preservation and access, has meant the evolution of software used for the management of digital content [25]. Open-source software such as digital asset management software (DAM) and institutional repository software [26] assist librarians in organizing digital items and providing a proper interface to their users. Integrated open-source software solutions such as Greenstone [27] are used for building entire digital libraries. Greenstone is destined particularly for universities, libraries, and other public service institutions. Aletheia [28] is a system for analysis and annotation of scanned documents, and started as a research project at the University of Salford Manchester. Reaching a new maturity of its development, it is now delivered as a Lite free version and a Pro version, full-featured, and with a commercial license. Another open-source project that started as a European research project is Ocropus/Ocropy [29]. Ocropus/Ocropy integrated multiple tools for OCR and document analysis. It integrates CLSTM [30], one of the first implementations of LSTM (Long-Short-Term-Memory) neural networks for character recognition.

3. Models of Digital Libraries

3.1. Global Level

At the global level, several countries and organizations established, over time, policies for digital sustainability. One such example is the International Federation of Library Associations who wrote a set of guidelines for correctly digitizing manuscripts and rare books [31]. Their purpose is to create a helping manual for any actor who wants to start a digitization process, including librarians, universities, researchers, and businesses.

In this section, we analyze some examples of digital libraries, including their goal, what they achieved, and if they had to face any difficulties in their implementations.

3.1.1. Google Books

Google Books (Figure 1), an international project for digitization, is a service created by Google in order to be able to index all the books available in bookstores and libraries. It started in 2002 when Google started experimenting in order to test its scanning technologies. The starting speed was about 450 pages/h in 2002; it then reached 6000 pages/h in 2015. The technology (patented by Google [32]) used two cameras and infrared light for scanning the books, as the pages were turned by a human operator. Efficient OCR algorithms were also developed, assuring the accuracy of the extracted text [33].

The number of books that have been scanned includes 25 million volumes, in 400 languages, from over 100 countries. It is stated that there are approximately 130 million distinct books in the world and Google Books has the intention of including them all in the collection [34]. Google Books can add knowledge through new research—for instance, linguists can analyze the evolution of the language of a certain culture by tracking the use of certain words or expressions over time.

However, the process is not flawless—besides some issues from the scanning process (a damaged paper which results in erroneous results from the OCR, unreadable pages, and blurry images) [35], Google has faced legal issues from various authors and publishers, with the most major debate being against the Association of American Publishers [36]. Google was sued and accused of “massive copyright infringement” [37], as many books that have been scanned were under copyright. After several years of lawsuits and numerous appeals, Google finally won the case, with its main argument being the fact that only snippets of copyrighted text are accessible to the public. The judge agreed that Google Books brings a great advantage for researchers, students, and teachers from all around the world while remaining respectful to the authors [38].

3.1.2. Europeana

Europeana [39] is a European Union initiative to provide digital access to cultural items throughout Europe via a network including all EU member states. Europeana (Figure 2) includes partnerships with thousands of museums, libraries, and archives to share cultural heritage; each partner provides access to digital items, adds a license, and publishes them with help from an aggregator [40]. It currently (i.e., 2020) provides access to 58 million digital items (books, music, and artwork) together with metadata for searching and filtering [41]. Europeana provides access to developers via an API (Application Programming Interface) and source code is available via an open-source license [42].

A subproject is Europeana Newspapers [43], providing access and metadata search, started in 2007, aggregating 18 million historical newspapers, available via Europeana.

Due to its openness, Europeana has been the target of academic evaluations that investigated criteria such as usability, availability, user-friendliness, and error rates [44]. These publications provide a foundation of criteria for evaluating digital libraries and for future standardization of methodology, tools, and measures for evaluation and comparison of different digital libraries.

Europeana is still considered to only cover about 15% of the possible digital items in Europe, with a major issue being that of intellectual property rights law in Europe [45] and the lack of institutional processes and interoperability [46]. Romania’s digital libraries were managed separately by different institutions with specific processes, making interoperability difficult. The introduction of the Romanian Digital Library [47] is expected to hasten the process of incorporation of Romanian digital items into Europeana.

3.2. National Level

Several countries have started the digitization process in order to preserve their patrimony. Below, we present some examples of national collections.

3.2.1. National Library of Norway

The National Library of Norway started the creation of its digital library in 2006, and they plan to offer digital access to the entire documentary heritage of the library. Currently, it consists of 104 thousand online books, and they estimate that the whole duration of the digitization process will last around 20 years [48]. The digitization process is based on three stages: document scanning, structure analysis, and post-processing. The collection is not limited to books; it also includes photographs, maps, magazines, newspapers, music, film, and TV programs [49]. The access to the content is open for the Norwegian citizens and its national importance is underlined by a large number of views: 51 million in 2012 [50].

3.2.2. Dutch National Library

In 2010, the Dutch National Library announced its plan to digitize all its printed publications, totaling a number of 730 million pages. They brought partners with relevant experience in the field, including Google and Proquest [51]. Currently, the collection includes thousands of databases with thousands of documents each, some of them open to the public while others may be accessed with a special paid account [52].

3.2.3. Digital Public Library of America (DPLA)

The Digital Public Library of America (DPLA) includes “35,259,924 images, texts, videos, and sounds”. Its main goal is to help people become aware of the American cultural heritage; access to the content is free, including books, photographs, maps, letters, or artwork [53]. The project received some critiques, especially for its redundant scope, similar to the one of Project Gutenberg—the oldest digital library which started in 1971 and is reaching over 60 thousand documents today [54].

3.3. Shadow Libraries

Over the last few decades [55], so-called “Shadow Libraries” have emerged. They stand at the border between being legal and illegal, offering a large variety of materials, free of access. Their resources are often pirated and stored on third-party websites, managing to dodge legislation. The need for these shadow libraries can be justified by the large number of students from all around the world who need access to academic material that they often cannot afford [56].

3.3.1. Sci-Hub

As mentioned on their website, Sci-Hub is self-entitled as “the first pirate website in the world to provide mass and public access to tens of millions of research papers”. It was founded in 2011 by Alexandra Elbakyan, a Kazakhstani Neural Engineering Graduate, and it provides free access to millions of books, regardless of their copyright status. Sci-Hub is able to copy documents that are restricted behind paywalls if they are requested by users. Elkabyan’s website has been described by Science Magazine as “an awe-inspiring act of altruism or a massive criminal enterprise, depending on whom you ask” [57]. Even if it had to face two lawsuits from publishers, Sci-Hub is praised by thousands of people, many of them claiming they would not have been able to write their thesis without the help of Sci-Hub [58].

3.3.2. Libgen

Library Genesis (shortly LibGen) is a massive online library, similar to Sci-Hub, created by Russian scientists. After losing the trial with Elsevier, the main website was closed, but mirrors stored on Russian servers resurfaced later. LibGen is currently blocked in a few countries, but with its database containing over 80 million documents, has found success all around the world [59].

3.3.3. Monoskop

In contrast to Sci-Hub or LibGen, Monoskop is described as a wiki for collaborative studies of culture [60]. Any user can edit articles and share new content, in a collaborative manner. It focuses on small collections which have been allegedly obtained through “donations”, regardless of their copyright status [61].

4. Case Study: Digitization in Romania

4.1. Normative Aspects and Statistics

The Romanian institutional framework for the digitization of the written patrimony started development in 2008, with the Order of the Minister of Culture number 2244/15 April 2008 regarding the establishment of the “Specialized Commissions for the digitization of the national cultural resources” [62] and the Decision 1676 from 10 December 2008 for the approval of the National program for the digitization of the national cultural resources and the creation of the Digital Library of Romania [63]. However, most of the legal norms in this field have an E.U. origin. The reference documents here are the “Recommendation of the European Commission on the digitization and online accessibility of the cultural material and digital preservation 2006/585/CE” on the 24th of August 2006 [64]; “The Conclusions of the EU Council C297 on the digitization and online accessibility of the cultural material”, published in the Official Journal of the EU on the 7th of December 2006 [65]; and the “Recommendation of the European Commission on the digitization and online accessibility of the cultural material and digital preservation 2011/711/UE” on 27 October 2011 [66].

Unfortunately, statistics about the digitization process of books in Romania are not encouraging, especially because the copyright legislation is uncertain, and the process is not cumbersome only if the authors of the books have been deceased for more than 70 years. The Ministry of Digitalization has the goal of digitalizing books in over 1000 libraries from cities and villages, yet it is interesting to analyze if Romanians are interested in digitalized books and whether the process makes sense to go further than just for preservation purposes [67]. According to a study conducted by Voxa, the biggest Romanian company distributing audiobooks and e-books, the digital book market in this country has a value of less than 1% of the total book sales [68].

4.2. Project Examples

4.2.1. Digital Library of Bucharest (Former Dacoromanica)

The Digital Library of Bucharest (formerly known as Dacoromanica) is a project developed by the Metropolitan Library of Bucharest, where old books from the Romanian patrimony have been digitized and made accessible to the public. The project is important for students, teachers, researchers, and all people interested in Romanian literature and culture [69].

However, the online portal suffers from technical difficulties caused by servers crashing and by the lack of permanent maintenance—subsequently, it has been offline for a significant period of time and with limited functionality [70]. Even if some of the technical issues has been solved in recent years, the user interface and user experience are outdated, affecting the platform’s popularity (Figure 3).

4.2.2. Memoria.ro

Memoria.ro is a digital library focusing on digitizing and storing “interviews, memoirs, oral history, books and pictures from the recent history of Romania, presenting events from the 20th century as reflected in the consciousness of those who lived them”. Memoria also contains full books in PDF format (whose copyright was obtained by the association), in Romanian, English, and French. It started in 2002 with 46 oral history interviews; it later reached hundreds of studies, books, and interviews, which can be sorted based on type, language, historical period, or region [71].

4.2.3. Medievalia

“Medievalia—fundamental texts of medieval Romanian culture” is a project that had the goal of identifying a priority corpus in the field of ancient Romanian literature and digitizing these manuscripts that contain valuable texts from the Romanian medieval era in order to provide a virtual research environment. The project had a duration of 10 months and managed to digitize a total number of 60 manuscripts regarding the old Romanian literature from the 15th to the 18th centuries. The manuscripts’ pages are scanned and available online, but only as images, as the manuscripts were written in old Cyrillic characters and are highly decorated, thus making the application of some form of OCR highly difficult. According to the Medievalia project, only a description with basic information (title, author, date, subject, and description) is available for the large public, as the studying of the scanned pages would be suitable mostly for historians and librarians. The platform offers just basic functionalities and is not user-friendly in terms of interface and intuitiveness (Figure 4a).

4.2.4. Byzantion Project

Byzantion Project is a research project similar to Medievalia, created by the Library of the Romanian Academy, which took place from 2012 to 2016. It managed to digitize a total number of 101 Byzantine manuscripts, dating back to the XIIth Century. The difficulty of the project comes from the fact that Byzantine writings are often highly decorated, including engravings and woodcuts [72]. Furthermore, each manuscript is different from the others, making the automation process challenging. The project managed to create a portal accessible to the public, where Byzantine manuscripts can be classified from a chronological, thematic, or geographical point of view [73]. Virtual canons have been created by studying all the images from the manuscripts and the well-known canons specific to the Orthodox iconography. The virtual canons are grouped in a data mining set, which can prove extremely useful for further research. As the project received funding from the Romanian Executive Agency for Higher Education, Research, Development and Innovation Funding (ro. Uefiscdi), the platform is better presented, easier to use, and more suitable for all actors which would be interested by the iconography subject (Figure 4b).

4.2.5. Lib2Life

The scientific project “Lib2Life—Revitalizing libraries and heritage through advanced technologies” has the purpose of conserving the Romanian cultural patrimony and promoting it through new, technologically enhanced means. It is launched as a collaboration between the University POLITEHNICA of Bucharest, The National Institute for Research and Development in Computer Science, and four national libraries.

5. Conclusions

In a world where we sometimes have to deal with unpredictable situations, digitization becomes essential as a bridge between people and culture. If we take into account the data related to the carbon emissions of printing books and we compare it with the production of electronic devices, we can understand that the latter can become sustainable if used efficiently and even lead to economic savings (e-books are usually cheaper than physical ones). As physical documents cannot be always accessed because of their lack of accessibility, the creation of digital content can raise the popularity of reading, increase the availability of knowledge all around the world, and decrease both illiteracy in its classical meaning and functional illiteracy. Furthermore, as sustainability policies implemented worldwide show, digitization can be essential for knowledge preservation, including rare books and manuscripts, preventing cultural loss.

In Romania, the digitization process is still not reaching the levels of other European projects, and measures are needed which include appropriate techniques, equipment, skilled professionals, and funding. Although the digital market for books is estimated today at just 1%, predictions suggest an increase to 10% in the following years [67]. In the future, we plan to continue more detailed research which will analyze the impact of digital libraries in Romania from an economic, social, cultural, and environmental point of view.