We show that the notion of bias in inductive concept learning can be quantified in a way that directly relates to learning performance, and that this quantitative theory of bias can provide guidance in the design of effective learning algorithms. We apply this idea by measuring some common language biases, including restriction to conjunctive concepts and conjunctive concepts with internal disj...

The ability to learn abstract relational concepts is fundamental to higher level cognition. In contrast to item-specific concepts (e.g. pictures containing trees versus pictures containing cars), abstract relational concepts are not bound to particular stimulus features, but instead involve the relationship between stimuli and therefore may be extrapolated to novel stimuli. Previous research in...

The effects of relevant concreteness on learning and transfer were investigated. Sixth grade students learned artificial instantiations of a simple mathematical concept. Some students were presented with instantiations that communicated concreteness relevant to the to-be-learned concept, while others learned generic instantiations involving abstract symbols. Results suggest that relevant concre...

Multi-modal models that learn semantic representations from both linguistic and perceptual input outperform language-only models on a range of evaluations, and better reflect human concept acquisition. Most perceptual input to such models corresponds to concrete noun concepts and the superiority of the multimodal approach has only been established when evaluating on such concepts. We therefore ...

We propose a novel approach to the machine learning of formal word sense, learned in interaction with human users using a new form of Relational Reinforcement Learning. The envisaged main application area of our framework is humanmachine communication, where a software agent or robot needs to understand concepts used by human users (e.g., in Natural Language Processing, HCI or Information Retri...

Experienced regularities in our perceptions and actions play important roles in grounding abstract concepts such as social status, time, and emotion. Might we similarly ground abstract spatial concepts in more experienced-based domains? The present experiment explores this possibility by implicitly priming abstract spatial terms (north, south, east, west) and then measuring participants' hand m...

Most of the modern formalisms used in Databases and Arti cial Intelligence for describing an application domain allow for using the notions of concept or class and relationship among con cepts There are basically two ways of using and describing classes In the rst one prescriptive approach the description formalism allows for expressing properties of the classes to be repre sented thus prescrib...

We show that the notion of bias in inductive concept learning can be quantified in a way that directly relates to learning performance, and that this quantitative theory of bias can provide guidance in the design of effective learning algorithms. We apply this idea by measuring some common language biases, including restriction to conjunctive concepts and conjunctive concepts with internal disj...