KBpedia v. 150 Released



CORALVILLE, IA (08/16/2017) -- KBpedia, a computable knowledge structure combining six major public knowledge bases, was released today in version 150. The purpose of KBpedia is to promote knowledge-based artificial intelligence (KBAI) and data interoperability. Version 150 has now added a major component to the knowledge graph enabling predicates to be split and organized according to attributes, external relations and representations. This split providea a framework to reason over all aspects of natural language, leading to more effective fact and relation extraction from knowledge bases.

A five-part series leading up to the release makes the point that most knowledge graphs focus on nouns and little attention has been given to properties or relations, especially as a classification of signs with key relevance to knowledge representation (KR). According to Mike Bergman, KBpedia's lead developer, "Actions drive the real changes in the world. Understanding them and their relationships, plus a more rigorous means for identifying and extracting them, should complete the integration of unstructured data with structured data."

According to Bergman the idea of categorizing predicates is not common in the knowledge representation space, but the writings of Charles Sanders Peirce were used to help provide guidance for how to organize this new KBpedia version.

The organization of relations into attributes (A:A), external relations (A:B) and representations (re:A) resulted in the addition of about 66 properties to KBpedia, now expressed in this version 1.50. These properties, in turn, have been mapped to about 2500 Wikidata properties, representing more than 90 percent of the property occurrences within that knowledge base. Via one or more properties, this mapping now extends KBpedia's coverage to about 30 million entities. KBpedia continues to have about 53,000 separate reference concepts.

The addition of these predicates also resulted in some fairly significant updates to the upper structure of KBpedia via the KBpedia Knowledge Ontology, or KKO. The new properties were also classed and categorized into the KKO node structure, a classic technique for being able to reason over predicates. More than 10% of the KKO knowledge graph was changed in version 1.50 to accommodate these changes.

In addition, the KBpedia hierarchy structure was further cleaned up to remove redundant subsumption assertions, leading to a cleaner and more understandable graph. Additional definitions were also added to the structure. In all, since the last release, Bergman estimates the entire KBpedia structure was re-built from scratch more than 100 times as all changes were incorporated, each time testing for logic and inconsistencies, including refinements to all of the existing 80 or so typologies in the system, especially the 30 "core" ones. Besides these changes, KBpedia's sponsor, Cognonto Corp., also established a separate Web site for the knowledge structure.

There is further documentation and an active knowledge graph on the KBpedia site. You can also run a demo showing how KBpedia information can inform a relatively simple tagger. The entire upper structure for KBpedia, KKO, is also available for download and inspection. A separate instruction version of KKO, which labels the major nodes according to the Peircean universal categories of Firstness, Secondness and Thirdness, is also helpful to learn more about the basic knowledge graph.

About KBpedia

The KBpedia knowledge structure combines six (6) public knowledge bases - Wikipedia, Wikidata, OpenCyc, GeoNames, DBpedia and UMBEL - into an integrated whole. These core KBs are supplemented with mappings to more than a score of additional leading vocabularies. The entire KBpedia structure is computable, meaning it can be reasoned over and logically sliced-and-diced to produce training sets and reference standards for machine learning and data interoperability. KBpedia greatly reduces the time and effort traditionally required for knowledge-based artificial intelligence (KBAI) tasks. KBpedia was first released in October 2016, though it has been under active development for more than six years. KBpedia is sponsored by Cognonto Corporation.

Press Contact

Mike Bergman, Cognonto Corp.