, , , , , , , ,

Technology is constantly changing society, either in a sustaining or disruptive way. From computers to the cloud, new possibilities have arisen to dismiss old habits. Moving towards the future implies perhaps not to give up entirely on these resources but to adapt them to enhance its performance.

Nowadays, the digital boom has empowered knowledge generation and information preservation, therefore efforts are being made to rescue information contained within ancient texts. Whether in large or small scale projects, digitation is growing strong, encompassing not only benefits as knowledge sharing, but also challenges that will play a main roll in the outcome.

Challenge 1: Completion of the task.

Analysing the Medici Archive Project [1], which comprehends the responsibility of digitizing around four million handwritten letters dating from the sixteenth to eighteenth centuries, 2 main challenges can be shown:

  • Size and nature of the collection.
  • Human Resource Management.

Depending on the nature of each text, i.e. handwritten letters, texts could be analysed automatically or manually, therefore having an impact in time and human capital. Furthermore, complexity of contents require more expertise of personnel in fields such as “paleography, language or history” becoming more difficult to gather the correct amount of people to complete the task. In addition, the Medici Archive contains texts in almost 5 different languages, thus increasing constraints in personnel skills and communications among them.

 Challenge 2: Assessing information and its limits.

Let’s for example take the Bookbinding example [2] to explain this challenge. In this article, they present the problem about inaccuracy of information related to information about bookbinding descriptions. This issue is given mainly due to a couple of reasons:

  • Absence of a well-established vocabulary and description system.
  • The randomly mix between words and drawings without a pattern.

Researches state that ignorance regarding how to define bookbinding itself was quite common, therefore in some cases uncertainty came embedded. In addition, as human beings, the source of information and its interpretation are subjective and as the complexity of the design and materials increased people have more difficulties to describe what they perceive, thus the possibility of human-error input also increase. For example, if the book have a lot of artistic work or it is too damaged to recognize its true origin.

They use main variables to portray uncertainty: location/position, size, value, texture, color, orientation, shape/form, color saturation, transparency/opacity, and sharpness/blurring. They chose blur to graphically show uncertainty.


Figure1.  Example of uncertainty of a sewing pattern.

Challenge 3: Assessing value of transformation.

Is it possible to say that by merely digitizing a text, the goal is achieved?  

The main target in the research made at the Hathi Trust Digital Library (international huge collaborative repository) [3] is to precisely promote dialogue to answer this question. They studied the impact that quality or the lack of it (error) might have in usefulness of books that were digitized at scale-projects, based on a model that stipulates the breach between the ideal digitation and the real one. Ideal meaning that digitation is made under process and standards.

Researches state that the result of a transformation of a book to digital code and algorithms is limited by digitation technologies, and they preserve signs of the process passed to create them. Results in a 10 million volume collection showed that the majority of books contain errors that do not compromise the reading, such as: thick/broken text, warped pages and obscured content, but there were extreme bad quality cases that jeopardize reliability of repositories.

In conclusion, technology tries to improve existing objects or tasks, but it is up to people to use them wisely and to recognize its limits. One should be able to distinguish when to use automation and when to make task manually. Moreover, the more complex a project can be, the more challenges it will face, therefore the use of process and standards can relieve the level of complication. Finally, the competition of a project does not always mean success, but it is achieved when the outcome encompasses requirements and it is useful.


[1] Allori, Lorenzo; Kaborycha, Lisa. Opening Aladdin’s cave or Pandora’s box? The challenges of crowdsourcing the Medici Archives. July 17, 2013. http://dh2013.unl.edu/abstracts/ab-312.html

[2] Campagnolo, Albert; Velios, Athanasios. Bidings of Uncertainty. Visualizing Uncertain and Imprecise Data in Automatically Generated Bookbinding Structure Diagrams. July 17, 2013.http://dh2013.unl.edu/abstracts/ab-187.html

[3] Conway, Paul. Surrogacy and Image Error: Transformations in the Value of Digitized Books. July 19, 2013. http://dh2013.unl.edu/abstracts/ab-363.html