The DOST Helpdesk is a Support Framework brings together a set of practical resources that guide contributors through the complete lifecycle of creating and contributing language datasets. The framework helps organizations, researchers, institutions, communities, and other contributors identify language resources, understand different types of language datasets, create and prepare datasets, and contribute them to the DOST ecosystem.

The resources explain what qualifies as a language dataset, including text, speech, translation, conversational, annotated, archival, cultural, OCR, ASR, TTS, transliteration, and other datasets, along with their potential applications in multilingual AI, research, public services, language preservation, and cultural documentation. They also provide guidance on assessing existing resources and transforming them into structured, clean, well-documented, privacy-conscious, and AI-ready language datasets.

The framework further covers the dataset onboarding process, including identification, Expression of Interest (EOI), review, preparation, metadata creation, upload, maintenance. The DOST Helpdesk acts as a support mechanism throughout this journey, providing guidance on dataset preparation, standardization, documentation, AI readiness, access to relevant tools and services, and post-publication support.

Overall, these resources can be used as a support framework for creating new language datasets as well as preparing existing language resources for AI use.