Explaining machine learning models with interactive natural language conversations using TalkToModel
Top Cited Papers
Open Access
- 27 July 2023
- journal article
- research article
- Published by Springer Nature in Nature Machine Intelligence
- Vol. 5 (8) , 873-883
Abstract
Practitioners increasingly use machine learning (ML) models, yet models have become more complex and harder to understand. To understand complex models, researchers have proposed techniques to explain model predictions. However, practitioners struggle to use explainability methods because they do not know which explanation to choose and how to interpret the explanation. Here we address the challenge of using explainability methods by proposing TalkToModel: an interactive dialogue system that explains ML models through natural language conversations. TalkToModel consists of three components: an adaptive dialogue engine that interprets natural language and generates meaningful responses; an execution component that constructs the explanations used in the conversation; and a conversational interface. In real-world evaluations, 73% of healthcare workers agreed they would use TalkToModel over existing systems for understanding a disease prediction model, and 85% of ML professionals agreed TalkToModel was easier to use, demonstrating that TalkToModel is highly effective for model explainability.Keywords
All Related Versions
Funding Information
- NSF | Directorate for Computer & Information Science & Engineering | Division of Information and Intelligent Systems (IIS-2040989, IIS-2046873, IIS-2008956, IIS-2008461, IIS-2040989, IIS-2008461, IIS-2040989, IIS-2040989, IIS-2046873, IIS-2008956)
- I was supported by a fellowship from the hasso plattner institute during the bulk of completing this work.
- Google, JP Morgan, Amazon, Harvard Data Science Initiative, D^3 institute at Harvard
This publication has 43 references indexed in Scilit:
- Prolific.ac—A subject pool for online experimentsJournal of Behavioral and Experimental Finance, 2018
- A Survey on Dialogue SystemsACM SIGKDD Explorations Newsletter, 2017
- Learning Certifiably Optimal Rule Lists2017
- Evaluating Semantic Parsing against a Simple Web-based Question Answering Model2017
- "Why Should I Trust You?"2016
- Interpretable Decision Sets2016
- Accurate intelligible models with pairwise interactions2013
- Developing Dialogue Managers from Limited Amounts of Data2012
- Psychological Aspects of Natural Language Use: Our Words, Our SelvesAnnual Review of Psychology, 2003
- Providing a unified account of definite noun phrases in discourse1983