Bell States and Negative Sentences in the Distributed Model of Meaning

被引：13

作者：

Preller, Anne ^{[1
]}

Sadrzadeh, Mehrnoosh ^{[2
]}

机构：

[1] Informatique, CNRS, LIRMM, Montpellier, France

[2] Univ Oxford, Comp Lab, Oxford, England

来源：

ELECTRONIC NOTES IN THEORETICAL COMPUTER SCIENCE | 2011年 / 270卷 / 02期

基金：

英国工程与自然科学研究理事会;

关键词：

Compact Closed Categories; Pregroups; Vector Spaces; Distributed Model of Meaning; Linguistics; Bell States;

D O I：

10.1016/j.entcs.2011.01.028

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

We use Bell states to provide compositional distributed meaning for negative sentences of English. The lexical meaning of each word of the sentence is a context vector obtained within the distributed model of meaning. The meaning of the sentence lives within the tensor space of the vector spaces of the words. Mathematically speaking, the meaning of a sentence is the image of a quantizing functor from the compact closed category that models the grammatical structure of the sentence (using Lambek Pregroups) to the compact closed category of finite dimensional vector spaces where the lexical meaning of the words are modeled. The meaning is computed via composing eta and epsilon maps that create Bell states and do substitution and as such allow the information to flow among the words within the sentence.

引用

页码：141 / 153

页数：13