Representing text as abstract images enables image classifiers to also simultaneously classify text

TL;DR

This paper presents a novel method to convert text data into abstract images, enabling the use of image classifiers for text comparison tasks like entity disambiguation, achieving high accuracy in patent inventor name matching.

Contribution

The paper introduces a new text-to-image representation technique that allows image classification models to be applied to text comparison problems, demonstrating its effectiveness in patent data.

Findings

High accuracy in inventor name disambiguation

Effective use of image classifiers on text comparison tasks

Potential applicability to broader NLP problems

Abstract

We introduce a novel method for converting text data into abstract image representations, which allows image-based processing techniques (e.g. image classification networks) to be applied to text-based comparison problems. We apply the technique to entity disambiguation of inventor names in US patents. The method involves converting text from each pairwise comparison between two inventor name records into a 2D RGB (stacked) image representation. We then train an image classification neural network to discriminate between such pairwise comparison images, and use the trained network to label each pair of records as either matched (same inventor) or non-matched (different inventors), obtaining highly accurate results. Our new text-to-image representation method could also be used more broadly for other NLP comparison problems, such as disambiguation of academic publications, or for problems that require simultaneous classification of both text and image datasets.