Personalized knowledge-based systems have not yet become widespread, despite their potential for valuable assistance in many daily tasks. This is due, in part, to the high cost of developing and maintaining customized knowledge bases. The construction of personal assistants as learning apprentices -interactive assistants that learn continually from their users -is one approach which could drama...

General-purpose knowledge bases (KBs) such as YAGO, Wikidata or the Google Knowledge Graph usually contain facts that have a high precision. In contrast, little is known about the recall of such KBs, and anecdotal evidence indicates that knowledge bases have a low recall on many topics. This project aims to develop techniques to evaluate and improve the recall of knowledge bases. We aim to use ...

In this paper, we investigate belief revision in possibilistic logic, which is a weighted logic proposed to deal with incomplete and uncertain information. Existing revision operators in possibilistic logic are restricted in the sense that the input information can only be a formula instead of a possibilistic knowledge base which is a set of weighted formulas. To break this restriction, we cons...

A poorly dcsigncd knowledge base can be as cryptic as an arbitrary program and just as difficult to maintain. Rcprescnting control knowledge abstractly, scparatcly from domain facts and relations, makes the design more transparent and explainable. A body of abstract control knowledge provides a generic framework for constructing knowledge bases for rclatcd problems in other domains and also pro...

The problem of maintaining a knowledge base is substantially concerned with keeping track of rules that share common wisdom. A knowledge representation is described in which a collection of rules that are based on common wisdom are represented as a single ‘item’. For items, maintenance hazards, caused by one item being partially hidden within another, still remain. ‘Objects’ are introduced as i...

Knowledge ‘decay’ is a measure of the degradation of knowledge integrity. In a unified knowledge representation, data, information and knowledge are all represented in a single formalism as “items”. A unified knowledge representation is extended here to include two measures of knowledge integrity. A rule of inference is defined on the unified knowledge representation that preserves the validity...

We develop a new methodology and tool for knowledge acquisition under uncertainty. A new knowledge representation called Bayesian Knowledge Bases provides a powerful key to our approach and is well-grounded in probability theory. In this paper, we demonstrate the ease and flexibility with which knowledge acquisition can be accomplished while ensuring the consistency of the knowledge base as dat...