Introduction
Heritage Crafts is a unique form of cultural heritage. It is a living heritage and includes both tangible and intangible elements. Artefacts and materials are tangible, and craftsmanship or knowledge about the process of making an artefact is intangible. The artisans are at the centre of this living legacy. They are the gatekeepers of traditional knowledge, and their hands reveal stories that are woven into the very fabric of their products; they are more than just tools. As the decline in the number of craft practitioners threatens the survival of many traditional craft forms, with the skilled artisans ageing and the younger generation opting for better livelihood options, preserving knowledge that lies with these artisans is essential to safeguard a nation's cultural heritage.
This paper discusses a novel method of digital preservation of Rajasthani (Indian) heritage crafts acquired through an initiative- Jodhpur City Knowledge and Innovation Foundation (JCKIF) promoted by the Indian Institute of Technology Jodhpur (IITJ). We use knowledge graphs to preserve artefacts, artisans, and their knowledge. Knowledge graphs are "a graph of data intended to accumulate and convey knowledge of the real world, whose nodes represent entities of interest and whose edges represent potentially different relations between these entities" (Hogan et al. 2021). Knowledge can be categorised into two types- declarative and procedural. Crafts, as a special form of heritage, contain both forms of knowledge. Artefacts and artisans are represented using declarative knowledge and craftsmanship as procedural knowledge. However, modelling procedural knowledge is a challenging task. It requires capturing a sequence of steps performed while making an artefact. This work presents a new method of preserving procedural knowledge by combining textual and video data. Videos of craft processes can be embedded with data in text format in the knowledge graph.
Handicrafts from Jodhpur, a city in the Indian state of Rajasthan, have been taken for this project. The five different types of crafts have been studied. Leather Mojari, Tie and Dye, Dabu Block Printing, Bone and Horn, and Durry Weaving. Leather Mojari is a lightweight embroidered footwear specific to this region of the country. Tie and Dye is a tie-resist dyeing technique done on fabrics. This craft is popularly known as "Bandhej" in the local language of this region. Dabu block printing is a mud-resist printing technique done on fabrics. Patterns are created using vegetable dyes and mud paste mixed with gum, lime, jaggery, and alum. Bone and Horn is a unique craft practice where utility and gift items are made of camel bones. Bones are carved or inlaid on wooden bases. It follows an ethical approach, using bones of only naturally dead animals. Durries are rugs and mainly weaved using cotton yarns. Apart from cotton, camel or sheep wool threads are used. Artisans use similar weaving methods to make yoga mats, coasters, table mats and handbags.
Furthermore, a digital museum of crafts has been built. It is a virtual reality space where artefacts are 3D-modelled and presented. There are galleries of all artisans, and their work has been showcased there. An attempt has been made to capture the reality of artisan's life and their workspaces. This paper also highlights how knowledge graphs can make this digital museum AI-empowered.
Existing Work
Knowledge graphs gained popularity with the advent of Google Knowledge Graphs in 2012. They are widely used in search engines, recommender systems, autonomous systems, and chatbots (Yse 2023). Lately, they have also found their application in the cultural heritage domain. Knowledge graph for Italian Cultural Heritage aimed at promoting tourism (Faralli, Lenzi, and Velardi 2022) , Knowledge graph for Intangible Cultural Heritage with intelligent question-answering system (Gu et al. 2022) , Multimodal knowledge graph with textual and image data for representing Chinese Intangible Cultural Heritage (Fan, Wang, and Hodel 2023) are a few examples of work done in this area.
However, most knowledge graphs are text-based, and little work has been done at the intersection of crafts, ontologies, and knowledge spaces. As part of the project Mingei , a project focusing on digital preservation and representation of heritage crafts, crafts ontology (CrO) has been developed (Zabulis et al. 2019) . In another project, a knowledge graph for preserving the Chinese silk weaving handicraft, Nanijing Yunjin, has been made using videos of the craft as primary data (Lu et al. 2023) .
The idea of museums in virtual reality has been there for a long time. With the advancements in digitisation methods and after witnessing the COVID-19 pandemic, many museums offer virtual tours. National Crafts Museum, India , British Museum, London , and Van Gogh Museum, Amsterdam are a few of them. Virtual reality enables wider public reach and heritage preservation at the same time. But, the amalgamation of AI in VR museums is still in the emerging phase.
Methodology
An ethnographic study of crafts has been done to build a holistic understanding of the crafts of Jodhpur. The 27 artisans from five different clusters were interviewed for this project. Artisan's stories about their lives, work, struggles, culture, and traditions are documented through audio recordings, videos, and textual data. Artefacts are 3D-modelled, and the process of making crafts is also recorded. A top-down approach is used for building knowledge graphs. An ontology of heritage crafts has been created first, followed by knowledge graphs. The CIDOC Conceptual Reference Model (CRM) is used as a base ontology for creating Indian_Crafts Ontology. It is both a practical and theoretical tool for integrating information in the cultural heritage domain.
Conclusion
We are building a digital museum driven by knowledge graphs using Indian Heritage encoding knowledge of artisans, artefacts and craftsmanship. It is an ongoing project aimed at digitally preserving heritage crafts and promoting artisans. We are currently working on how to embed knowledge graphs in the digital museum. In the future this method of preserving (in)tangible heritage can be enhanced to paintings and folk music.