1. A Milestone Worth Celebrating

The Arahanism Publications corpus has now been made openly accessible through two complementary digital archival platforms: Zenodo and Hugging Face Datasets.

This marks an important milestone in our ongoing effort to preserve, organize, and transmit the original liberation teaching of Gautama, the Supreme Arahant, the Sammāsambuddha.

Through this dual-channel archival strategy, our publications are now available both for long-term scholarly preservation and for modern computational research and artificial intelligence applications.

Together, these platforms establish two complementary pathways for the preservation and dissemination of Arahanism publications: one serving the enduring needs of scholarly documentation, and the other opening our written works to the developing possibilities of artificial intelligence and machine-assisted knowledge discovery.

This achievement is not simply about making books available online. It represents a commitment to ensuring that the dhamma teachings, research, translations, and writings developed through Arahanism can remain accessible to present and future generations.

We are pleased to mark this milestone and to share these resources with the wider human community.

2. Knowledge Should Be Shared Freely

At the heart of this initiative lies a simple conviction: knowledge should be preserved and shared freely for the benefit of humanity.

This conviction is particularly important when it comes to the Dhamma—the original liberation teaching of Gautama, the Supreme Arahant, the Sammāsambuddha.

The Dhamma is not merely a collection of philosophical ideas, historical records, or religious literature. It is a path of practice leading towards liberation from rāga, dosa, and moha—the underlying defilements that bind beings to the conditioned world.

Such a teaching deserves to be preserved, studied, understood, and transmitted across generations.

Our intention in making Arahanism publications openly accessible is therefore not limited to expanding the readership of our books. It is also to contribute to the wider body of human knowledge by making our research and written materials available to anyone who wishes to study, examine, reference, or build upon them.

We believe that the preservation of knowledge should not depend exclusively on commercial distribution or the limitations of physical publications. Modern digital technology offers opportunities to make written knowledge accessible on a much broader scale.

Accordingly, our digital corpus is made available under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence, allowing others to share and adapt the materials in accordance with its attribution requirements.

This reflects an aspiration that extends beyond our own institution: that knowledge may continue to circulate, that research may inspire further research, and that future generations may have access to the teachings and ideas preserved by those who came before them.

3. Two Channels, One Purpose: Zenodo and Hugging Face

To achieve both long-term scholarly preservation and accessibility within the developing AI ecosystem, we have adopted a dual-channel archival strategy using two complementary platforms: Zenodo and Hugging Face.

Although both platforms make our publications openly accessible, they serve different purposes.

Zenodo: Permanent Scholarly Archival

Zenodo is an open-access research repository developed by CERN and supported through the European OpenAIRE infrastructure. It provides researchers, scholars, and institutions with a platform for preserving and sharing research outputs, publications, datasets, and other scholarly materials.

By depositing Arahanism publications on Zenodo, we establish formal scholarly records with Digital Object Identifiers (DOIs). These identifiers allow individual works to be cited, referenced, and located through stable academic records.

This provides an important foundation for the long-term preservation and scholarly recognition of our publications. Researchers and readers can identify specific works, reference them in their own research, and access the original materials without depending exclusively on commercial publishing platforms.

Zenodo therefore serves as our principal channel for scholarly documentation and archival preservation.

Hugging Face: Open Access for Artificial Intelligence

Hugging Face serves a different but equally important purpose.

It has become a major platform within the open-source artificial intelligence community, providing infrastructure for sharing datasets, machine-learning models, and related research resources.

By publishing the Arahanism Publications corpus as an openly accessible dataset on Hugging Face, we make our written works available in structured formats suitable for computational processing and automated data ingestion.

This creates opportunities for AI researchers, developers, and independent research communities to incorporate our publications into their own projects, including Retrieval-Augmented Generation (RAG) systems, knowledge repositories, semantic search applications, and other AI-assisted research tools.

Rather than limiting access to conventional reading formats, we are making the underlying textual content available in formats that computers can process directly.

Hugging Face therefore serves as our principal channel for machine-readable distribution and AI research accessibility.

One Purpose, Two Complementary Channels

These two platforms represent different dimensions of the same commitment.

Zenodo helps ensure that our publications have identifiable and citable scholarly records. Hugging Face makes their textual content accessible to the wider computational research community.

Together, they establish an infrastructure through which Arahanism publications can be preserved, discovered, studied, referenced, and potentially incorporated into future research and knowledge systems.

Our objective is straightforward: to make these teachings and research materials accessible not only to readers today, but also to the technological and scholarly environments through which knowledge will be explored in the future.

4. Making the Dhamma Accessible to AI

One distinctive aspect of this initiative is that our publications have not simply been uploaded as conventional digital documents. Their textual content has also been prepared in machine-readable formats to facilitate automated processing by modern artificial intelligence systems.

This distinction is important.

A conventional PDF or printed book is primarily designed for human reading. Although computers can extract text from many PDFs, the underlying structure may not always be convenient for automated processing, indexing, or retrieval.

Machine-readable formats provide a more direct way for computational systems to access and interpret written content.

Markdown: Structured and Readable Text

Our publications are made available in Markdown format, a lightweight text format that preserves essential document structures, including headings, paragraphs, lists, and other textual elements.

Markdown offers two important advantages: it remains easily readable by humans while also being straightforward for software applications to process.

This makes it particularly useful for organizing extensive collections of written teachings, translations, commentaries, and research materials.

JSONL: Structured Data for Automated Processing

We also use JSONL (JSON Lines), a format in which each line contains an independent JSON record.

This structure allows individual textual entries or document segments to be processed independently, making the corpus suitable for automated data pipelines, dataset loaders, indexing systems, and AI research applications.

For example, a developer building a Dhamma knowledge repository or a Retrieval-Augmented Generation (RAG) application can process the records and organize them for subsequent retrieval and reference.

Rather than requiring every document to be manually opened and processed, structured records make large collections of textual knowledge considerably easier to handle programmatically.

UTF-8: Preserving Pāḷi and Its Original Terminology

Particular attention has also been given to preserving Pāḷi terminology and diacritical marks.

Pāḷi contains characters that are essential to accurate transliteration and scholarly communication, including:

ā, ī, ū, ṅ, ñ, ṭ, ḍ, ṇ, ḷ, and other specialized characters.

The corpus uses UTF-8 text encoding, which supports these characters and allows them to be represented consistently across compatible software systems and digital platforms.

This is especially important for Arahanist and Buddhist textual research, where seemingly small differences in spelling and diacritics can distinguish different Pāḷi words and forms.

Preserving these characters helps maintain the textual integrity of our publications when they are processed, indexed, searched, or incorporated into other compatible digital research environments.

Opening the Door to AI-Assisted Dhamma Research

By combining structured Markdown documents, JSONL records, and UTF-8 encoding, we are making the Arahanism Publications corpus more accessible to modern computational research.

Researchers and developers can potentially use these materials to:

• Build searchable digital repositories of Arahanism publications.
• Develop Retrieval-Augmented Generation (RAG) applications for accessing and referencing the teachings.
• Conduct computational analysis of Pāḷi terminology, textual relationships, and doctrinal concepts.
• Integrate the publications into AI-assisted research and educational tools.
• Develop specialized knowledge bases for studying Gautama’s liberation teaching.

Importantly, making a dataset openly accessible does not automatically mean that every AI model will ingest it or incorporate it into its training. Actual utilization depends on the developers, researchers, and systems that choose to access and process the materials.

Nevertheless, by preparing the corpus for machine ingestion, we remove an important technical barrier and make these publications available to a much wider range of future applications.

We are entering an era in which artificial intelligence will increasingly participate in the discovery, organization, and transmission of human knowledge.

We believe that the original liberation teaching of Gautama should not be excluded from this developing knowledge environment.