• Publications
  • Influence
Graded Entailment for Compositional Distributional Semantics
TLDR
The main theorem shows that entailment strength lifts compositionally to the sentence level, giving a lower bound on sentence entailment. Expand
Interacting Conceptual Spaces I : Grammatical Composition of Concepts
TLDR
This work introduces the category of convex relations as a new setting for categorical compositional semantics, emphasizing the convex structure important to conceptual space applications, and shows how to construct conceptual spaces for various types such as nouns, adjectives and verbs. Expand
Modelling the interplay of metaphor and emotion through multitask learning
TLDR
This paper investigates the relationship between metaphor and emotion within a computational framework, by proposing the first joint model of these phenomena, and demonstrates that metaphor identification and emotion prediction mutually benefit from joint learning. Expand
Interacting Conceptual Spaces
TLDR
This work introduces the category of convex relations as a new setting for categorical compositional semantics, emphasizing the convex structure important to conceptual space applications, and shows how conceptual spaces for composite types such as adjectives and verbs can be constructed. Expand
Towards logical negation for compositional distributional semantics
TLDR
This paper gives some steps towards providing this operator of logical negation, modelling it as a version of projection onto the subspace orthogonal to a word, and gives a small demonstration of the operators performance in a sentence entailment task. Expand
Hierarchical conceptual spaces for concept combination
TLDR
It is shown that the conjunctive model includes linear combination as a special case, and that the more general model can account for non-compositional behaviours such as overextension, non-commutativity, preservation of necessity and impossibility of attributes and to some extent, attribute loss or emergence. Expand
Compositional Hyponymy with Positive Operators
TLDR
The word representations built can be understood within a compositional distributional semantic framework, providing methods for composing words to form phrase and sentence level representations. Expand
A Compositional Explanation of the 'Pet Fish' Phenomenon
TLDR
It is shown here how the `pet fish' phenomenon is naturally accommodated within a compositional distributional model of meaning, which describes the meaning of a composite concept by accounting for interaction between its constituents via their grammatical roles. Expand
Modelling hyponymy for DisCoCat
In natural language processing words are commonly represented as vectors. However, vector representations do not intrinsically incorporate the hierarchical relationships that obtain between manyExpand
Cats climb entails mammals move: preserving hyponymy in compositional distributional semantics
TLDR
This paper proposes that psd matrices for verbs, adjectives, and other functional words be lifted to completely positive maps that match their grammatical type, and gives a number of proposals for the structure of Compr, based on spiders, cups and caps, and generates a range of composition rules. Expand
...
1
2
3
4
...