ThaiCoGen: Knowledge Graph Generation Enriching Suvarnabhumi’s Historical Insights

Historical tourism documentation includes guidebooks, pamphlets, and digital resources, which are collections of materials that promote the exploration of historical sites, events, and cultural heritage. These materials aim to offer insights into the history and importance of particular locations, people, and items. Within these documents, nominal historical terminologies possess valuable information, serving as significant entities that can contribute to our understanding of history. This study introduces a method for knowledge graph (KG) generation through Thai co-occurrence network analysis, called ThaiCoGen. This work improves KG generation by using three important methods: (1) term frequency-inverse document frequency (TF-IDF) to find important terms and entities; (2) Thai co-occurrence network text analysis (ThaiCoNet) to connect semantic relationships; and (3) adding to KGs using linked open data from encyclopedias. Our contributions are the generation of KGs using extracted high-n-gram entities and the highlighting of semantic enrichment. Our contributions include (1) generating KGs by identifying the extracted height n-gram entities and (2) highlighting semantic enrichment. A case study was conducted on the Thai encyclopedia LST.org in the domain of the Suvarnabhumi Silk Road and demonstrated the efficacy of the ThaiCoGen methodology. Despite the substantial temporal gap between historical records from the Suvarnabhumi Silk Road era (100–700 BE) and contemporary tourist documentation from the modern era (2500 BE), ThaiCoGen successfully integrates these sources into enriched KGs. The discussion of this paper intends to elaborate on the historically accurate portrayal of the Suvarnabhumi Silk Road from the enriched KG, especially Thailand's tourism industry. Finally, this research provides the open KGs for Thailand's cultural heritage in order to build the robustness of semantic application development.