Design of Knowledge Graph in Manufacturing Services Discovery

Abstract

Abstract Knowledge graph networks powering web search and chatbot agents have shown immense popularity. This paper discusses the first steps towards building a knowledge graph for manufacturing services discoverability. Due to the lack of a unified widely adopted schema for structured data in the manufacturing services domain as well as the limitations of existing relational database schemas to represent manufacturing service definitions, there does not exist a unified schema that connects manufacturing resources service descriptions and actual manufacturing service business entities. This gap severely limits the automated discoverability of manufacturing service business organizations. This paper designs a knowledge graph covering over 8,000+ manufacturers, the manufacturing services they provide and corresponding linkage with manufacturing service definitions available from Wikidata. In addition, this work also proposes extensions to Schema.org to assist small business manufacturers to contain embedded search engine optimization (SEO) tags for search and discovery through web search engines. Such vocabulary extensions are critical to the rapid identification and real-time capability assessment particularly when the service providers themselves are responsible for updating tags. A wider scale enhancement of manufacturing specific vocabulary extensions to schema.org can tremendously benefit small and medium scale manufacturers. This paper concludes with the additional work that must be done for a comprehensive addition to manufacturing service graph that spans the entire manufacturing knowledge base.