Chapter 20: Knowledge-Graphs

by Mayank KejriwalUniversity of Southern California

Since the introduction of the Google Knowledge Graph in the early 2010s, web search and advertising have both undergone profound shifts. A Knowledge Graph (KG) is a graph-based way of providing entities, relations, events, facts, and other aspects of knowledge to machines. By representing such objects formally, data science and Artificial Intelligence (AI) algorithms are able to ‘understand’ sophisticated queries and comprehend complex user needs, including recommendations, allowing for new possibilities to be unlocked. Beyond the Google Knowledge Graph and generic web search, however, the application of KGs has been less prominent across the industry. In the commercial domain, product-related e-commerce sectors have been a major area of focus, as embodied in efforts such as the Amazon Product Graph, but the same cannot be said of service-related sectors such as tourism. Therefore, this is an opportunity waiting to be exploited. In this book chapter, some of the fundamentals of KGs will be explained, along with details on how a KG can be built and used within the tourism domain. Practical guidelines on implementation will also be discussed, drawing on existing technological tools, as well as a set of hints and tips.