Interpreting Contrastive Embeddings in Specific Domains with Fuzzy Rules

TL;DR

This paper introduces a fuzzy rule-based method to interpret CLIP embeddings in specific domains like clinical reports and film reviews, enhancing understanding of feature importance and domain adaptation.

Contribution

It presents a novel approach combining fuzzy rules with CLIP embeddings to improve interpretability and domain-specific analysis of text representations.

Findings

Effective mapping of features to CLIP space in two domains

Analysis of rule-based associations and feature importance

Discussion of limitations and potential improvements

Abstract

Free-style text is still one of the common ways in which data is registered in real environments, like legal procedures and medical records. Because of that, there have been significant efforts in the area of natural language processing to convert these texts into a structured format, which standard machine learning methods can then exploit. One of the most popular methods to embed text into a vectorial representation is the Contrastive Language-Image Pre-training model (CLIP), which was trained using both image and text. Although the representations computed by CLIP have been very successful in zero-show and few-shot learning problems, they still have problems when applied to a particular domain. In this work, we use a fuzzy rule-based classification system along with some standard text procedure techniques to map some of our features of interest to the space created by a CLIP model. Then, we discuss the rules and associations obtained and the importance of each feature considered. We apply this approach in two different data domains, clinical reports and film reviews, and compare the results obtained individually and when considering both. Finally, we discuss the limitations of this approach and how it could be further improved.