Calibration of Machine Reading Systems at Scale

Abstract

In typical machine learning systems, an estimate of the probability of the prediction is used to assess the system's confidence in the prediction. This confidence measure is usually uncalibrated; i.e. the system's confidence in the prediction does not match the true probability of the predicted output. In this paper, we present an investigation into calibrating open setting machine reading systems such as open-domain question answering and claim verification systems. We show that calibrating such complex systems which contain a discrete retrieval and deep reading components is challenging and current calibration techniques fail to scale to these settings. We propose simple extensions to existing calibration approaches that allow us to adapt these callibrators to these settings. Our experimental results reveal that the joint callibration of the retriever and the reader outperforms the reader calibrator by a significant margin. We also show that the callibrator can be useful for selective prediction, e.g., when question answering systems are posed with unanswerable or out-ofthe-training distribution questions.

Revisiting Calibration for Question Answering

“…While selective question answering offers a way of measuring calibration, the scale of confidence values are not considered since abstention can be effective as long as correct predictions have higher confidence than wrong ones, regardless of the absolute scales. A recent work (Dhuliawala et al, 2022) also explored calibration for retriever-reader style ODQA setup, mainly focusing on how to combine information from the retriever and reader components. While in this paper we only focused on span-extraction QA, Jiang et al (2021) focused on generation-style QA.…”