THE BRITISH LIBRARY

Digital scholarship blog

3 posts from September 2019

20 September 2019

Labbers of the world unite to write a book in 1 week through a Book Sprint

Posted by Mahendra Mahey Manager of BL Labs.

I can't believe it's been a year since people from national, state, regional, university libraries (as well as a few galleries, archives and museums) met in London to attend the first global 'Library Labs' event at the British Library on 13th and 14th of September 2018. These 'Labs' are increasingly found in cultural heritage and academic institutions around the world and offer a space for their users to experiment and innovate on-site and on-line with their own (and others') digitised and born digital collections and data.

We had over 70 people from 43 institutions and 20 countries attend the London event and it was really wonderful, with a very full programme. There was a palpable sense of excitement and willingness to want to share experiences, build new professional relationships and witness the birth of a new international 'Labs' community. Through the event, we were able to understand more about the digital 'Labs' landscape around the world from the results of Library Labs survey. For example, we learned that many institutions were in the process of planning a 'Lab', many wanted to learn more about how to set them up, maintain and sustain them and learn the lessons from those that had already done it. About half of the attendees in London had already set up Labs in their organisations and wanted to share their experiences with other professionals so that they could build better Labs and help others so they didn't have to reinvent the wheel to save time and precious resources.

Growing an international Cultural Heritage Labs community
Some of the presenters from the first Building Library Labs Event at the
British Library, London, UK on 13-14 September 2019

The event was a mixture of presentations and lightning talks, stories of how labs are developing, parallel discussion groups and debates, many of which were videoed. At the end of the event, the collaborative document we had created contained over 60 edited pages of notes, together with a folder of other useful documents and presentations. It was concluded that it would be wonderful to come together to perhaps convert these shared experiences into a useful book/guide, perhaps through a Book Sprint. A Book Sprint is where up to 15 people come together for a week, and with minimal distractions work together to create a book. Each day when the participants sleep, a team of illustrators and editors transform their content for the next day remotely. The week ends having created a book! A great idea for busy people! We felt it was a nice fit for the Labs community we work in or want to create, which are largely based on a 'mindset' of experimentation, taking risks and being prepared to learn from your mistakes. I started to research how it might be possible to hold such a Book Sprint by talking to the Book Sprint company that has had over 20 years experience organising and running these book creation events.

Collectively as a group we decided that we would continue to build the Labs community and establish a mailing list. Clemens Neudecker wrote an excellent blog post about the event.

Zoom meeting Building Library LabsA screen grab from a virtual zoom meeting of the building Labs community

Subsequently, we held various meetings from October 2018 through to February 2019 (some virtual and some face to face) and agreed to hold our next global Labs meeting at the Royal Danish Library in Copenhagen, Denmark on 4-5 March 2019, again with an action packed programme with the help of Katrine Gasser and her team at kbtechlab. Directly after that event, some of us participated in a pre-conference workshop as part of Digital Humanities Nordic 2019, DHN-Labs - Digital Humanities and the National and University Libraries and Archives (in the Nordic and Baltic Countries) on the 6 March 2019.

Royal Danish Library, Copenhagen, Denmark

Royal Danish Library, Copenhagen, Denmark where the second
Building Library Labs event was held between 4-5 March, 2019

Over 50 people attended the 2-day event in Copenhagen, although similar to the previous event in London, this time we agreed we would hold it under Chatham House rule (an idea from Kirsty Lingstadt from the University of Edinburgh) which many of us found was very liberating.

Again, we managed to produce over 60 pages of notes and collect other relevant and helpful information. It was even more abundantly clear at the end of this event that we would definitely need to find a way for some of us to come together to write a book through the Book Sprint methodology previously proposed.

A very kind and generous offer of exploring funding from her institution was made by Milena Dobreva-McPherson Associate Professor Library and Information Studies at University College London Qatar. Abigail Potter from the Library of Congress Labs also kindly suggested that she and her team may be able to hold the next global Labs meeting in Washington between 4-6 May, 2020 in the USA.

Myself and Milena met in Qatar at the first Musuem's and Big Data conference in Qatar organised by her colleague Georgios Papaioannou Associate Professor of Museum Studies, in May 2019. We formulated a proposal to UCL Qatar (funded by the Qatar Foundation) which was successful. Milena also managed to also obtain funding from the University of Qatar. There has also been support from the British Library Labs, the Library of Congress Labs, Book Sprint Ltd, who agreed to donate half of the Book Sprint fee to run the event and finally Qatar National Library.

What was important from the outset was that the digital version of the book should be made FREELY available on the web to reuse, in line with the spirit and ethos of the group.

Milena also managed to secure funding for research assistants Somia Salim and Fidelity Phiri to help create a global directory of organisations which are doing Labs style things or might want to. They have also helped out and are helping at various Labs style events including the Book Sprint.

From the first building library labs event in September 2018 to the present day there have been various events where the work of this community has been mentioned. Here is a small sample:

In July 2019, we released an open invitation to apply to be part of the Book Sprint and received some fantastic entries. We would like to thank everyone that sent an application and we would like to reassure everyone that they can still contribute to the community even if they were not chosen on this occasion.

We can now finally announce who will be attending the Book Sprint...drum droll...:

  1. Abigail Potter, Senior Innovation Specialist with the Library of Congress Digital Innovation Lab. She tweets at @opba.
  2. Aisha Al Abdulla, Section Head of the Digital Repository and Archives at Qatar University Library.
  3. Caleb Derven, Head of Technical and Digital Services at the University of Limerick with overall responsibility for strategy and operations related to collections, electronic resources and library systems. He tweets at @calebderven.
  4. Ditte Laursen, Head of Department, The Royal Library Denmark responsible for the acquisition of digitally born cultural heritage materials, long-term preservation of digital heritage collections, and access to digital cultural heritage collections. She tweets at @DitteDla.
  5. Gustavo Candela, Associate Professor at the University of Alicante and member of the Research and Development department at The Biblioteca Virtual Miguel de Cervantes. He tweets at @gus_candela.
  6. Katrine Gasser, Section Head of IT at The Royal Library Denmark managing a team of 40 IT experts in programming, networking and research. She tweets at @blackat_ and kbtechlab
  7. Kristy Kokegei, Director of Public Engagement at the History Trust of South Australia who oversees the organisation’s public programming, digital engagement, marketing, learning and education programs across 4 State Government funded museums and supporting and enabling 350 community museums and historical societies across South Australia. She tweets at @KristyKokegei and @SAGLAMLab.
  8. Lotte Wilms, Digital Scholarship advisor managing the KB Research Lab and Digital Humanities in libraries advocate, co-chair for the LIBER working group Digital Humanities and a board member of the IMPACT Centre of Competence. She tweets at @Lottewilms.
  9. Mahendra Mahey, Manager of British Library Labs (BL Labs), an Andrew W. Mellon foundation and British Library funded initiative supporting and inspiring the use of its data in innovative ways with scholars, artists, entrepreneurs, educators and innovators through competitions, awards and other engagement activities. He tweets at @BL_Labs and @mahendra_mahey.
  10. Milena Dobreva-McPherson, Associate Professor Library and Information Studies at UCL Qatar with international experience of working in Bulgaria, Scotland and Malta. She tweets at @Milena_Dobreva.
  11. Paula Bray, DX Lab Leader at the State Library of NSW and responsible for developing and promoting an innovation lab utilising emerging and existing web technologies to deliver new ways to explore the Library’s collections and its data. She tweets at @paulabray #dxlab @statelibrarynsw
  12. Sally Chambers, Digital Humanities Research Coordinator at Ghent Centre for Digital Humanities, Ghent University, Belgium and National Coordinator for DARIAH, the Digital Research Infrastructure for the Arts and Humanities in Belgium. She tweets at @schambers3, @GhentCDH and @KBRbe
  13. Sarah Ames, Digital Scholarship Librarian at the National Library of Scotland, responsible for developing a Digital Scholarship Service and launching the Data Foundry. She tweets at @semames1.
  14. Sophie-Carolin Wagner, Co-Founder of RIAT Research Institute for Art and Technology, Co-Editor of the Journal for Research Cultures and Project Manager of ONB Labs at the Austrian National Library.
  15. Stefan Karner, Technical Lead of the ONB Labs at the Austrian National Library, providing access to diverse data and metadata sources within the library, developing a platform for users of the digital library to create and share annotations and other user generated data with each other and the public.
  16. Armin Straube, Teaching Fellow in Library and Information Studies at UCL Qatar. He is an archivist with work experience in data curation, digital preservation and web archiving and tweets at @ArminStraube.

Laia Ros Gasch will be facilitating the Book Sprint and has 10 years of experience as a cultural producer working all over the world with all kinds of groups. Laia speaks English, French, Spanish and Catalan. 

More detailed biographies are available here.

WE NEED YOUR HELP!

We all realise how incredibly lucky and privileged we are to be chosen. However, we want to hear from those of you who are interested in this area. What do you think we should be writing about, who should it be for, what style of writing should we use? Please HELP us by completing this questionnaire by Monday 23 September at 0600 BST! We will consider your thoughts and opinions seriously when we sit down to write the book on Monday morning in Doha in Qatar.

We would also like to get your help when we will be disseminating information about how to get hold of the book on social media, and at various events around the world, especially to coincide with International Open Access week 2019 (21-27 October 2019). Planned activities in 2019-2020 include:

We plan to run a 'Read Sprint' in the near future to review the Book and perhaps create an improved version. We know what we will produce next week won't be perfect!

We have plans to ensure that the book is published on a interactive platform so that it becomes a 'living' book, so that others can add chapters, make amendments, enhancements and add new case studies. We will be making announcements about this soon after the book has been completed.

On a personal note, I feel incredibly grateful, lucky and privileged to have been involved at the very start of this journey. I also feel daunted to be part of the Book Sprint but excited too!

I really want us to create a useful handbook to help cultural heritage organisations build better innovation labs which are often strapped for resources and need help. I have a strong desire that our ‘Book’ will genuinely help and inspire galleries, libraries, archives, museums, universities and other cultural heritage organisations to learn and benefit from those of us who can talk honestly about and share our experiences. I want to share the risks we have taken, mistakes we have made, provide realistic lessons and give sensible advice about what we have learned over the many years in setting up, maintaining and sustaining innovation labs. I believe this approach could mean it may prevent many institutions from having to re-invent the wheel and save them time, money and resources too.

The people in this community have a passionate desire to create something useful and meaningful that will help all of us be better at our jobs and build better innovation labs for the benefit of all our users. Hopefully, we will be following the principles of kindness, generously sharing and understanding and having empathy for the contexts in which we work. In short we hope it sincerely makes a difference and prove that sharing and kindness really can change things.

Now that I have written this, I realise I have done it again, I have written too much! However, I am glad I have written the story of how we got here. What I realise is what a busy year it’s been for everyone and particularly for people in this community, it’s amazing what we have achieved and I want to thank everyone who has played an active role, no matter how small. Let’s hope it continues to grow.

Monday morning, fifteen of us have got to write a book, gulp!

14 September 2019

BL Labs Awards 2019: enter before 2100 on Sunday 29th September! (deadline extended)

We have extended our deadline for our BL Labs Awards to 21:00 (BST) on Sunday 29th September, submit your entry here. If you have already entered, you don't have to resubmit, however, we are happy to receive updated entries too.

The BL Labs Awards formally recognises outstanding and innovative work that has been created using the British Library’s digital collections and data.

Submit your entry, and help us spread the word to all interested parties!

This year, BL Labs is commending work in four key areas:

  • Research - A project or activity that shows the development of new knowledge, research methods, or tools.
  • Commercial - An activity that delivers or develops commercial value in the context of new products, tools, or services that build on, incorporate, or enhance the Library's digital content.
  • Artistic - An artistic or creative endeavour that inspires, stimulates, amazes and provokes.
  • Teaching / Learning - Quality learning experiences created for learners of any age and ability that use the Library's digital content.

After the submission deadline of 21:00 (BST) on Sunday 29th September for entering the BL Labs Awards has passed, the entries will be shortlisted. Selected shortlisted entrants will be notified via email by midnight BST on Thursday 10th October 2019. 

A prize of £500 will be awarded to the winner and £100 to the runner up in each Awards category at the BL Labs Symposium on 11th November 2019 at the British Library, St Pancras, London.

The talent of the BL Labs Awards winners and runners up over the last four years has led to the production of a remarkable and varied collection of innovative projects. In 2018, the Awards commended work in four main categories – Research, Artistic, Commercial and Teaching & Learning:

Photo collage

  • Research category Award (2018) winner: The Delius Catalogue of Works: the production of a comprehensive catalogue of works by the composer Delius, based on research using (and integrated with) the BL’s Archives and Manuscripts Catalogue by Joanna Bullivant, Daniel Grimley, David Lewis and Kevin Page from Oxford University’s Music department.
  • Artistic Award (2018) winner: Another Intelligence Sings (AI Sings): an interactive, immersive sound-art installation, which uses AI to transform environmental sound recordings from the BL’s sound archive by Amanda Baum, Rose Leahy and Rob Walker independent artists and experience designers.
  • Commercial Award (2018) winner: Fashion presentation for London Fashion Week by Nabil Nayal: the Library collection - a fashion collection inspired by digitised Elizabethan-era manuscripts from the BL, culminating in several fashion shows/events/commissions including one at the BL in London.
  • Teaching and Learning (2018) winner: Pocket Miscellanies: ten online pocket-book ‘zines’ featuring images taken from the BL digitised medieval manuscripts collection by Jonah Coman, PhD student at Glasgow School of Art.

For further information about BL Labs or our Awards, please contact us at labs@bl.uk.

Posted by Mahendra Mahey, Manager of of British Library Labs.

13 September 2019

Results of the RASM2019 Competition on Recognition of Historical Arabic Scientific Manuscripts

This blog post is by Dr Adi Keinan-Schoonbaert, Digital Curator for Asian and African Collections, British Library. She's on Twitter as @BL_AdiKS.

 

Earlier this year, the British Library in collaboration with PRImA Research Lab and the Alan Turing Institute launched a competition on the Recognition of Historical Arabic Scientific Manuscripts, or in short, RASM2019. This competition was held in the context of the 15th International Conference on Document Analysis and Recognition (ICDAR2019). It was the second competition of this type, following RASM2018 which took place in 2018.

The Library has an extensive collection of Arabic manuscripts, comprising of almost 15,000 works. We have been digitising several hundred manuscripts as part of the British Library/Qatar Foundation Partnership, making them available on Qatar Digital Library. A natural next-step would be the creation of machine-readable content from scanned images, for enhanced search and whole new avenues of research.

Running a competition helps us identify software providers and tool developers, as well as introduce us to the specific challenges that pattern recognition systems face when dealing with historic, handwritten materials. For this year’s competition we provided a ground truth set of 120 images and associated XML files: 20 pages to be used to train text recognition systems to automatically identify Arabic script, and a 100 pages to evaluate the training.

Aside from providing larger training and evaluation sets, for this year’s competition we’ve added an extra challenge – marginalia. Notes written in the margins are often less consistent and less coherent than main blocks of text, and can go in different directions. The competition set out three different challenges: page segmentation, text line detection and Optical Character Recognition (OCR). Tackling marginalia was a bonus challenge!

We had just one submission for this year’s competition – RDI Company, Cairo University, who previously participated in 2018 and did very well. RDI submitted three different methods, and participated in two challenges: text line segmentation and OCR. When evaluating the results, PRImA compared established systems used in industry and academia – Tesseract 4.0, ABBYY FineReader Engine 12 (FRE12), and Google Cloud Vision API – to RDI’s submitted methods. The evaluation approach was the same as last year’s, with PRImA evaluating page analysis and recognition methods using different evaluation metrics, in order to gain an insight into the algorithms.

 

Results

Challenge 1 - Page Layout Analysis

The first challenge was set out to identify regions in a page, and find out where blocks of text are located on the page. RDI did not participate in this challenge, therefore an analysis was made only on common industry software mentioned above. The results can be seen in the chart below:

Chart showing RASM2019 page segmentation results
Chart showing RASM2019 page segmentation results

 

Google did relatively well here, and the results are quite similar to last year’s. Despite dealing with the more challenging marginalia text, Google’s previous accuracy score (70.6%) has gone down only very slightly to a still impressive 69.3%.

Example image showing Google’s page segmentation
Example image showing Google’s page segmentation

 

Tesseract 4 and FRE12 scored very similarly, with Tesseract decreasing from last year’s 54.5%. Interestingly, FRE12’s performance on text blocks including marginalia (42.5%) was better than last year’s FRE11 performance without marginalia, scoring at 40.9%. Analysis showed that Tesseract and FRE often misclassified text areas as illustrations, with FRE doing better than Tesseract in this regard.

 

Challenge 2 - Text Line Segmentation

The second challenge looked into segmenting text into distinct text lines. RDI submitted three methods for this challenge, all of which returned the text lines of the main text block (as they did not wish to participate in the marginalia challenge). Results were then compared with Tesseract and FineReader, and are reflected below:

Chart showing RASM2019 text line segmentation results
Chart showing RASM2019 text line segmentation results

 

RDI did very well with its three methods, with an accuracy level ranging between 76.6% and 77.6%. However, despite not attempting to segments marginalia text lines, their methods did not perform as well as last year’s method (with 81.6% accuracy). Their methods did seem to detect some marginalia, though very little overall, as seen in the screenshot below.

Example image showing RDI’s text line segmentation results
Example image showing RDI’s text line segmentation results

 

Tesseract and FineReader again scored lower than RDI, both with decreasing accuracy compared to RASM2018’s results (Tesseract 4 with 44.2%, FRE11 with 43.2%). This is due to the additional marginalia challenge. The Google method does not detect text lines, therefore the Text Line chart above does not include their results.

 

Challenge 3 - OCR Accuracy

The third and last challenge was all about text recognition, tackling the correct identification of characters and words in the text. Evaluation for this challenge was conducted four times: 1) on the whole page, including marginalia, 2) only on main blocks of text, excluding marginalia, 3) using the original texts, and 4) using normalised texts. Text normalisation was performed for both ground truth and OCR results, due to the historic nature of the material, occasional unusual spelling, and use/lack of diacritics. All methods performed slightly better when not tested on marginalia; accuracy rates are demonstrated in the charts below:

Chart showing OCR accuracy results, for main text body only (normalised, no marginalia)
Chart showing OCR accuracy results, for main text body only (normalised, no marginalia)
 
Chart showing OCR accuracy results for all text regions (normalised, with marginalia)
Chart showing OCR accuracy results for all text regions (normalised, with marginalia)

 

It is evident that there are minor differences in the character accuracies for the three RDI methods, with RDI2 performing slightly better than the others. When comparing the OCR accuracy between texts with and without marginalia, there are slightly higher success rates for the latter, though the difference is not significant. This means that tested methods performed on the marginalia almost as well as they did on the main text, which is encouraging.

Comparing RASM2018’s results, RDI’s results are good but not as good as last year (with 85.44% accuracy), likely to be a result of adding marginalia to the recognition challenge. Google performed very well too, considering they did not specifically train or optimised for this competition. Tesseract’s results went down from 30.45% to 25.13%, and FineReader Engine 12 performed better than its previous version FRE11, going up from 12.23% to 17.53% accuracy. However, it is still very low, as handwritten texts are not part of their target material.

 

Further Thoughts

RDI-Corporation has its own historical Arabic handwritten and typewritten OCR system, which has been built using different historical manuscripts. Its methods have done well, given the very challenging nature of the documents. Neither Tesseract nor ABBYY FineReader produce usable results, but that’s not surprising since they are both optimised for printed texts, and target contemporary material and not historical manuscripts.

As next steps, we would like to test these materials with Transkribus, which produced promising results for early printed Indian texts (see e.g. Tom Derrick’s blog post – stay tuned for some even more impressive results!), and potentially Kraken as well. All ground truth will be released through the Library’s future Open Access repository (now in testing phase), as well as through the website of IMPACT Centre for Competence. Watch this space for any developments!