A Sea of Words: An In-Depth Analysis of Anchors for Text Data

TL;DR

This paper provides a theoretical analysis of the Anchors interpretability method for text classification, examining its behavior across different models and offering insights into its selection mechanism, especially for neural networks.

Contribution

It offers the first formal analysis of Anchors for text data, including explicit results for models with TF-IDF vectorization and insights into neural network explanations.

Findings

Anchors effectively identify influential words in text classification.

Theoretical insights apply to models like linear classifiers and rule-based systems.

Empirical evidence shows Anchors select words with high partial derivatives for neural networks.

Abstract

Anchors (Ribeiro et al., 2018) is a post-hoc, rule-based interpretability method. For text data, it proposes to explain a decision by highlighting a small set of words (an anchor) such that the model to explain has similar outputs when they are present in a document. In this paper, we present the first theoretical analysis of Anchors, considering that the search for the best anchor is exhaustive. After formalizing the algorithm for text classification, we present explicit results on different classes of models when the vectorization step is TF-IDF, and words are replaced by a fixed out-of-dictionary token when removed. Our inquiry covers models such as elementary if-then rules and linear classifiers. We then leverage this analysis to gain insights on the behavior of Anchors for any differentiable classifiers. For neural networks, we empirically show that the words corresponding to the highest partial derivatives of the model with respect to the input, reweighted by the inverse document frequencies, are selected by Anchors.