Predicting Entity Popularity to Improve Spoken Entity Recognition by Virtual Assistants

TL;DR

This paper presents a method that leverages historical user interaction data to predict emerging entity popularity, enhancing spoken entity recognition in virtual assistants and reducing errors by 20% on trending entities.

Contribution

It introduces a novel approach to forecast emerging entity popularity and integrates it into the ASR system to improve recognition accuracy for trending entities.

Findings

20% relative error reduction on emerging entity utterances

Improved recognition of trending entities without degrading overall system performance

Effective use of historical user interaction data for entity prediction

Abstract

We focus on improving the effectiveness of a Virtual Assistant (VA) in recognizing emerging entities in spoken queries. We introduce a method that uses historical user interactions to forecast which entities will gain in popularity and become trending, and it subsequently integrates the predictions within the Automated Speech Recognition (ASR) component of the VA. Experiments show that our proposed approach results in a 20% relative reduction in errors on emerging entity name utterances without degrading the overall recognition quality of the system.