# Interacting with Explanations through Critiquing

arXiv: https://arxiv.org/abs/2005.11067 · HTML: https://mlsanity.com/paper/2005.11067/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 May 2020.

## Built on by (most cited first, 9 of 9)

- [Leveraging Explanations in Interactive Machine Learning: An Overview](https://mlsanity.com/paper/2207.14526/) (2022)
- [Rationalization through Concepts](https://mlsanity.com/paper/2105.04837/) (2021)
- [Editable User Profiles for Controllable Text Recommendation](https://mlsanity.com/paper/2304.04250/) (2023)
- [A survey on improving NLP models with human explanations](https://mlsanity.com/paper/2204.08892/) (2022)
- [Fast Multi-Step Critiquing for VAE-based Recommender Systems](https://mlsanity.com/paper/2105.00774/) (2021)
- [Self-Supervised Bot Play for Conversational Recommendation with Justifications](https://mlsanity.com/paper/2112.05197/) (2021)
- [Assistive Recipe Editing through Critiquing](https://mlsanity.com/paper/2205.02454/) (2022)
- [Recommending Burgers based on Pizza Preferences: Addressing Data Sparsity with a Product of Experts](https://mlsanity.com/paper/2104.12822/) (2021)
- [Positive and Negative Critiquing for VAE-based Recommenders](https://mlsanity.com/paper/2204.02162/) (2022)

## Abstract (arXiv)

Using personalized explanations to support recommendations has been shown to increase trust and perceived quality. However, to actually obtain better recommendations, there needs to be a means for users to modify the recommendation criteria by interacting with the explanation. We present a novel technique using aspect markers that learns to generate personalized explanations of recommendations from review texts, and we show that human users significantly prefer these explanations over those produced by state-of-the-art techniques. Our work's most important innovation is that it allows users to react to a recommendation by critiquing the textual explanation: removing (symmetrically adding) certain aspects they dislike or that are no longer relevant (symmetrically that are of interest). The system updates its user model and the resulting recommendations according to the critique. This is based on a novel unsupervised critiquing method for single- and multi-step critiquing with textual explanations. Experiments on two real-world datasets show that our system is the first to achieve good performance in adapting to the preferences expressed in multi-step critiquing.