Jump to content

2024:Program/Wikisource Loves Manuscripts: Lessons & Horizons

From Wikimania
View on Commons

Session title: Wikisource Loves Manuscripts: Lessons & Horizons

Session type: Poster
Track: GLAM
Language: en

This discussion will explore the "Wikisource Loves Manuscripts" (WiLMa) project, a collaborative effort to preserve and revitalize handwritten manuscripts on Wikimedia projects.

We will delve into the project's multifaceted approach, encompassing manuscript digitization, community-driven transcription through Wikisource proofread-a-thons, and the utilization of AI-powered tools like Transkribus for enhanced Optical Character Recognition (OCR) in under-resourced languages.

We will share learnings from the pilot project in Indonesia focusing on its three regions and languages and how the resources produced during the pilot helped convene a network of learning partners interested in working with manuscripts on Wikisource.

Description

[edit | edit source]

The "Wikisource Loves Manuscripts" (WiLMa) project is a call-to-action and a project to support the digitization of manuscripts on Wikisource through a collaborative approach. This session will bring together wikimedians, partners like Transkribus, and WMF staff to share the project's innovative strategies.

We'll delve into how WiLMa leverages Wikisource, a platform enabling a global volunteer network to transcribe vast amounts of manuscript material. We'll also explore the learnings from the project's Indonesian pilot, which focused on digitizing and transcribing texts across three regions and languages. We will explore WiLMa partnerships with organizations like UNESCO and the British Library. Panelists will discuss how these partnerships have played a crucial role in supporting the project's goals. For instance, the collaboration with UNESCO has helped to raise awareness of the importance of cultural heritage preservation, while the partnership with the British Library has facilitated the digital repatriation of cultural heritage materials.

The session will explore WiLMa's use of AI to address challenges in under-resourced languages. We'll showcase the project's integration of Transkribus, an AI-powered tool, with Wikisource. This integration allows training of OCR/HTR models for languages like Balinese and Javanese.

The session will also introduce the WiLMa Learning Partners Network, designed to empower participants with the skills and resources necessary to undertake their own manuscript digitization projects such as the digitization guidelines created during the pilot in Indonesia. The first cohort of the network saw 22 participants from 15 languages and ended in March 2024. We will share what are some of the community-driven projects being undertaken by the network members and other experts.

How does your session relate to the event theme, Collaboration of the Open?

The WiLMa project hinges on "Collaboration of the Open" at every step. The project brings together Wikimedia affiliates (like WMID), Partners (like PPIM, UNESCO and the British Library), individual contributors, and Wikimedia Foundation for a shared goal.

While volunteer contributors contribute to Wikisource to tackle the transcription effort, while partners like Transkribus provide the AI tools to help overcome tech barriers such as unavailability of handwritten text recognition (HTR) technology. The project also fosters collaboration within communities through the Learning Partners Network, empowering them to preserve their own manuscripts. Finally, partnerships with UNESCO and the British Library raise awareness and provide resources (such as opening up access to Javanese manuscripts from the Library), making WiLMa's mission of safeguarding handwritten manuscripts a true team effort.

What is the experience level needed for the audience for your session?

Everyone can participate in this session

Resources

[edit | edit source]

Speakers

[edit | edit source]
  • Satdeep Gill
I work with the Wikimedia Foundation as a Senior Program Officer in the Culture and Heritage team, where I'm passionate about supporting Wikisource and language inclusion within the Wikimedia movement.
In my volunteer capacity, I have been a Wikimedia contributor since 2009 mainly contributing to Punjabi Wikipedia, Punjabi Wikisource and Wikimedia Commons.