EXS: Explainable Search Using Local Model Agnostic Interpretability
By Jaspreet Singh, Avishek Anand. 2018
117 citations recorded by the source corpus.
Retrieval models in information retrieval are used to rank documents for typically under-specified queries. Today machine learning is used to learn retrieval models from click logs and/or relevance judgments that maximizes an objective correlated with user satisfaction. As these models become increasingly powerful and sophisticated, they also become harder to understand. Consequently, it is hard for to identify artifacts in training, data specific biases and intents from a complex trained model like neural rankers even if trained purely on text features. EXS is a search system designed specifically to provide its users with insight into the following questions: "What is the intent of the query according to the ranker?'', "Why is this document ranked higher than another?'' and "Why is this document relevant to the query?''. EXS uses a version of a popular posthoc explanation method for classifiers -- LIME, adapted specifically to answer these questions. We show how such a system can effectively help a user understand the results of neural rankers and highlight areas of improvement.
Subfields: Artificial Intelligence
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Salvatore Ruggieri
- Dino Pedreschi
- Richard Socher
- Caiming Xiong
- Riccardo Guidotti
- Fosca Giannotti
- Nazneen Rajani
- Anna Monreale
Citations in the corpus
- Explainable Recommendation: A Survey and New Perspectives (2018, background): “It is also worth noting that explainability perspectives have also been integrated into other intelligent systems beyond recommendation, such as explainable search (Singh and Anand, 2019), question answering (Zhao et al.”
- BERTology Meets Biology: Interpreting Attention in Protein Language Models (2020, background): “Examples include approaches that fit a linear classifier to a model’s prediction on perturbations of the input [4, 61, 72].”
- Meaningful Explanations of Black Box AI Decision Systems (2019, background): “A few more recent meth-ods are model-agnostic, such as LIME (Ribeiro, Singh, and Guestrin 2016; Singh and Anand 2018).”
- Explaining and Improving Model Behavior with k Nearest Neighbor Representations (2020, background): “It has been extended to identifying artifacts and biases in information retrieval (Singh and Anand 2018).”