Apple is sponsoring the forty-third Special Interest Group on Information Retrieval (SIGIR) Conference, which will be held virtually from July 25 to 30. SIGIR is a leading international forum focused on presenting new research in the informational retrieval field.

Learn more about SIGIR

Conference Accepted Paper

Predicting Entity Popularity to Improve Spoken Entity Recognition by Virtual Assistants

Christophe Van Gysel, Manos Tsagkias, Ernest Pusateri, Ilya Oparin

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 twenty percent relative reduction in errors on emerging entity name utterances without degrading the overall recognition quality of the system.

Related readings and updates.

A Discriminative Entity Aware Language Model for Virtual Assistants

High-quality automatic speech recognition (ASR) is essential for virtual assistants (VAs) to work well. However, ASR often performs poorly on VA requests containing named entities. In this work, we start from the observation that many ASR errors on named entities are inconsistent with real-world knowledge. We extend previous discriminative n-gram language modeling approaches to incorporate real-world knowledge from a Knowledge Graph (KG), using…
See paper details

Predicting Entity Popularity to Improve Spoken Entity Recognition by Virtual Assistants

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…
See paper details