# Evaluating vector-space models of analogy

arXiv: https://arxiv.org/abs/1705.04416 · HTML: https://mlsanity.com/paper/1705.04416/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 May 2017.

## Built on by (most cited first, 6 of 6)

- [Fluid Transformers and Creative Analogies: Exploring Large Language Models' Capacity for Augmenting Cross-Domain Analogical Creativity](https://mlsanity.com/paper/2302.12832/) (2023)
- [A Neural Approach for Detecting Morphological Analogies](https://mlsanity.com/paper/2108.03945/) (2021)
- [Theoretical foundations and limits of word embeddings: what types of meaning can they capture?](https://mlsanity.com/paper/2107.10413/) (2021)
- [Tackling Morphological Analogies Using Deep Learning -- Extended Version](https://mlsanity.com/paper/2111.05147/) (2021)
- [Tversky Neural Networks: Psychologically Plausible Deep Learning with Differentiable Tversky Similarity](https://mlsanity.com/paper/2506.11035/) (2025)
- [Social World Knowledge: Modeling and Applications](https://mlsanity.com/paper/2306.16299/) (2023)

## Abstract (arXiv)

Vector-space representations provide geometric tools for reasoning about the similarity of a set of objects and their relationships. Recent machine learning methods for deriving vector-space embeddings of words (e.g., word2vec) have achieved considerable success in natural language processing. These vector spaces have also been shown to exhibit a surprising capacity to capture verbal analogies, with similar results for natural images, giving new life to a classic model of analogies as parallelograms that was first proposed by cognitive scientists. We evaluate the parallelogram model of analogy as applied to modern word embeddings, providing a detailed analysis of the extent to which this approach captures human relational similarity judgments in a large benchmark dataset. We find that that some semantic relationships are better captured than others. We then provide evidence for deeper limitations of the parallelogram model based on the intrinsic geometric constraints of vector spaces, paralleling classic results for first-order similarity.