Partners Look to Scale ‘Chomsky Knowledge Graph’
Noam Chomsky, the outspoken and prolific linguist, cognitive scientist, philosopher and social critic, is the subject of the first public Knowledge Graph, a compendium of Chomsky’s scholarly and other writings and interviews along with the commentaries and new reports Chomsky’s vast body of work has generated.
Franz Inc., a semantic graph database technology vendor based in Oakland, Calif., and the Semantic Web Co. have teamed to produce the first semantic knowledge graph of a public figure, assembling what they say is a “knowledge model of Chomsky’s brain [and] how his ideas evolved,” said Fred Davis, who heads the knowledge graph project.
“Our goal is to make Chomsky’s work searchable in the context of topics and concepts, readable in excerpts and easily available to journalists, scientists, technologists, students, philosophers and historians as well as the general public,” Davis added.
Knowledge graphs represent a broad and deep subject framework that connects different types of information in a systematic way, using taxonomies to describe ontology models that define categories and the links among concepts, data and entities. Upon that foundation custom models can be built, the Chomsky Knowledge Graph being the first.
Among the advances represented by knowledge graphs is their ability to encode knowledge arranged in networks of nodes and links rather than traditional tables of rows and columns.
That approach is touted as allowing data scientists and automation tools such as natural language processing to build what the partners call a “semantic network of facts” about entities. The capability could then be used for data integration, semantic queries and data analysis.
The exercise also could serve as a template for addressing broader data management issues in industries ranging from financial services and energy to health care and life sciences, proponents said. “The application of knowledge graphs to public figures, such as Noam Chomsky, will offer a unique opportunity to link concepts and ideas to form new ideas and possible solutions,” added Andreas Blumauer, Semantic Web’s CEO and founder.
The partners said they are currently prototyping their platform consisting of taxonomy and ontology models used to organize and identify the connections within large data sets ranging from books and movies to newspaper and scholarly articles. They expect to release a full-blown knowledge graph platform by the end of 2019.
The initiative builds on Google’s (NASDAQ: GOOGL) pioneering work that used a graph database to organize information and link it to related web sites and sources. Built on top of the graph database, the knowledge graph transformed search from “strings” of key words to “things,” or concepts based on context.
“Context requires connections, and graphs – as complex systems – offer the highest level of context,” Jim Webber, chief scientist at graph database vendor Neo4j, wrote in a guest column for Datanami earlier this year. “We’ve seen that these graphs are actually hungry for new connections and new data, which in turn creates the opportunity for data scientists to evolve and refine the algorithms that operate upon the graph.”
The Chomsky Knowledge Graph seeks to take this approach a step further by leveraging the graph database’s “nodes and links” structure as the foundation of a knowledge graph representing all that can be known about a given subject or entity. “You gather everything you can around those entities” and use semantic search for much improved queries, said Jans Aasman, CEO of Franz Inc.
Despite the breadth and types of data connected in knowledge graphs, Aasman added that “technical scalability is not an issue,” noting that Franz’s semantic graph technologies are already used in data-intensive sectors like health care.
Among the reasons Chomsky was selected for the first knowledge graph is that the polymath is considered the most widely cited author in the world, Davis noted in an interview.
The Chomsky Knowledge Graph will be hosted by the Internet Archive, the world’s largest digital lending library.