This text discusses the process of extracting knowledge from raw texts and creating a knowledge graph using tools like GCP, GPT, and Wikidata. A knowledge graph is a graph that represents relationships between entities, concepts, and facts in a specific domain. The text highlights the importance of taxonomies in enhancing the power of a knowledge graph. It also mentions the challenges in navigating and building a knowledge graph, including the need for programming skills, linguistics, and domain knowledge. The text emphasizes the role of natural language processing (NLP) in the creation of knowledge graphs and the use of domain-specific NLP tools. It concludes by mentioning the author’s semi-automatic NLP pipeline for generating augmented knowledge graphs from both English and Japanese raw texts.
| Signal | Change | 10y horizon | Driving force |
|---|---|---|---|
| Extracting knowledge with GCP, GPT, and Wikidata | From manual extraction to semi-automatic | More efficient and accurate knowledge extraction | Advancements in AI and natural language processing |
| Use of knowledge graph for complex relationships | From manual analysis to visualization | Enhanced visualization of complex relationships | Need for efficient decision-making and problem-solving |
| No-code tools for navigating knowledge graph | From complex navigation to user-friendly | Easier navigation and exploration of knowledge | Simplification of technology for wider adoption |
| Challenges in building a knowledge graph | From uncomplicated to complex | Improved tools and techniques for graph creation | Need for more advanced programming and domain skills |
| Augmented knowledge graph from raw texts | From limited information to augmented data | Enriched knowledge graph with additional entities | Integration of external data sources and technologies |