Projects of Knowledge Graphs

mitigating risk, tracking news events, and increasing transport automation, amongst (many) others. We now discuss the main industries in which enterprise knowledge graphs have been deployed.

Web search. Web search engines have traditionally focused on matching a query string with sub-strings in web  documents. The Google Knowledge Graph rather promoted a paradigm of “things not strings” – analogous to  semantic search – where the search engine would now try to identify the entities that a particular search may be  expressing interest in. The knowledge graph itself describes these entities and how they interrelate. One of the main  userfacing applications of the Google Knowledge Graph is the “Knowledge Panel”, which presents a pane on the right- hand side of (some)  search results describing the principal entity that the search appears to be seeking, including some  images, attribute–value pairs, and a list of related entities that users also search for.

Commerce. Enterprise knowledge graphs have also been announced by companies that are principally concerned with  selling or renting goods and services. A prominent example of such a knowledge graph is that used by Amazon, which describes the products on sale in their online marketplace. One of the main stated goals of this knowledge  graph is to enable more advanced (semantic) search features for products, as well as to improve product  recommendations to users of its online marketplace. Another knowledge graph for commerce was announced by eBay, which encodes product descriptions and shopping behaviour patterns, and is used to power conversational agents that aid users to find relevant products through a natural language interface. Airbnb have also described a knowledge graph that encodes accommodation for rent, places, events, experiences, neighbourhoods, users, tags, etc., on top of  which a taxonomic schema is defined. This knowledge graph is used to offer potential clients recommendations of  attractions, events, and activities available in the neighbourhood of a particular home for rent. Uber have similarly announced a knowledge graph focused on food and restaurants for their “Uber Eats” delivery service. The goals are again to offer semantic search features and recommendations to users who are uncertain precisely what kind of food they are  looking for. 

Social networks. Enterprise knowledge graphs have also emerged in the context of social networking services. Facebook  have gathered together a knowledge graph describing not only social data about users, but also the entities they are interested in, including celebrities, places, movies, music, etc., in order to connect people, understand their interests, and  provide recommendations. LinkedIn announced a knowledge graph containing users, jobs, skills, companies, places, schools, etc., on top of which a taxonomic schema is defined. The knowledge graph is used to provide multilingual  translations of important concepts, to improve targetted advertising, to provide advanced features for job search and  people search, and likewise to provide recommendations matching jobs to people (and vice versa). Another knowledge  graph has been created by Pinterest, describing users and their interests, the latter being organised into a taxonomy.  The main use-cases for the knowledge graph are to aid users to more easily find content of interest to them, as well as to enhance revenue through targetted advertisements.

Finance. The financial sector has also seen deployment of enterprise knowledge graphs. Amongst these, Bloomberg has proposed a knowledge graph that powers financial data analytics, including sentiment analysis for companies based on current news reports and tweets, a question answering service, as well as detecting emerging events that may affect  stock values. Thompson Reuters (Refinitiv) have likewise announced a knowledge graph encoding “the financial  ecosystem” of people, organisations, equity instruments, industry classifications, joint ventures and alliances, supply  chains, etc., using a taxonomic schema to organise these entities. Some of the applications they mention for the  knowledge graph include supply chain monitoring, risk assessment, and investment research. Knowledge graphs have also  been explored in academic settings with Banca d’Italia, using rule-based reasoning to determine, for example,  the percentage of ownership of a company by various stakeholders. Other companies exploring financial knowledge  graphs include Accenture, Capital One, Wells Fargo, amongst others.

Other industries. Enterprises have also been actively developing knowledge graphs to enable novel applications in a  variety of other industries, including: health-care, where IBM are exploring use-cases for drug discovery and  information extraction from package inserts, while AstraZeneca are using a knowledge graph to advance  genomics research and disease understanding; transport, where Bosch are exploring a knowledge graph of scenes and  locations for driving automation ; oil & gas, where Maana are using knowledge graphs to perform data  integration for risk mitigation regarding oil wells and drilling; and more besides 

Some Knowledge Graphs sources