A Sea of Words: An In-Depth Analysis of Anchors for Text Data
This work provides foundational insights into how Anchors works for text classification, benefiting researchers in interpretable machine learning, but it is incremental as it builds on an existing method without introducing new paradigms.
The paper presents the first theoretical analysis of the Anchors interpretability method for text data, formalizing the algorithm and deriving explicit results for models like linear classifiers and if-then rules, and empirically showing that for neural networks, Anchors selects words based on high partial derivatives reweighted by inverse document frequencies.
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.