SleepQA: A health coaching dataset on sleep for extractive question answering

I Bojic; QC Ong; M Thakkar; E Kamran; IY Le Shua; JRE Pang; J Chen; V Nayak; S Joty; J Car

SleepQA: A health coaching dataset on sleep for extractive question answering

I Bojic, QC Ong, M Thakkar, E Kamran, IY Le Shua, JRE Pang, J Chen, V Nayak, S Joty, J Car

Research output: Other contribution

Abstract

Question Answering (QA) systems can support health coaches in facilitating clients' lifestyle behavior changes (e.g., in adopting healthy sleep habits). In this paper, we design a domain-specific QA pipeline for sleep coaching. To this end, we release SleepQA, a dataset created from 7,005 passages comprising 4,250 training examples with single annotations and 750 examples with 5-way annotations1. We fine-Tuned different domain-specific BERT models on our dataset and perform extensive automatic and human evaluation of the resulting end-To-end QA pipeline. Comparisons of our pipeline with baseline show improvements in domain-specific natural language processing on realworld questions. We hope that this dataset will lead to wider research interest in this important health domain.

Original language	Undefined/Unknown
Number of pages	19
Volume	193
Publication status	Published - 1 Jan 2022

Cite this

@misc{07e10721743a42249b2384bc109b92f9,

title = "SleepQA: A health coaching dataset on sleep for extractive question answering",

abstract = "Question Answering (QA) systems can support health coaches in facilitating clients' lifestyle behavior changes (e.g., in adopting healthy sleep habits). In this paper, we design a domain-specific QA pipeline for sleep coaching. To this end, we release SleepQA, a dataset created from 7,005 passages comprising 4,250 training examples with single annotations and 750 examples with 5-way annotations1. We fine-Tuned different domain-specific BERT models on our dataset and perform extensive automatic and human evaluation of the resulting end-To-end QA pipeline. Comparisons of our pipeline with baseline show improvements in domain-specific natural language processing on realworld questions. We hope that this dataset will lead to wider research interest in this important health domain.",

author = "I Bojic and QC Ong and M Thakkar and E Kamran and {Le Shua}, IY and JRE Pang and J Chen and V Nayak and S Joty and J Car",

year = "2022",

month = jan,

day = "1",

language = "Undefined/Unknown",

volume = "193",

type = "Other",

}

TY - GEN

T1 - SleepQA: A health coaching dataset on sleep for extractive question answering

AU - Bojic, I

AU - Ong, QC

AU - Thakkar, M

AU - Kamran, E

AU - Le Shua, IY

AU - Pang, JRE

AU - Chen, J

AU - Nayak, V

AU - Joty, S

AU - Car, J

PY - 2022/1/1

Y1 - 2022/1/1

N2 - Question Answering (QA) systems can support health coaches in facilitating clients' lifestyle behavior changes (e.g., in adopting healthy sleep habits). In this paper, we design a domain-specific QA pipeline for sleep coaching. To this end, we release SleepQA, a dataset created from 7,005 passages comprising 4,250 training examples with single annotations and 750 examples with 5-way annotations1. We fine-Tuned different domain-specific BERT models on our dataset and perform extensive automatic and human evaluation of the resulting end-To-end QA pipeline. Comparisons of our pipeline with baseline show improvements in domain-specific natural language processing on realworld questions. We hope that this dataset will lead to wider research interest in this important health domain.

AB - Question Answering (QA) systems can support health coaches in facilitating clients' lifestyle behavior changes (e.g., in adopting healthy sleep habits). In this paper, we design a domain-specific QA pipeline for sleep coaching. To this end, we release SleepQA, a dataset created from 7,005 passages comprising 4,250 training examples with single annotations and 750 examples with 5-way annotations1. We fine-Tuned different domain-specific BERT models on our dataset and perform extensive automatic and human evaluation of the resulting end-To-end QA pipeline. Comparisons of our pipeline with baseline show improvements in domain-specific natural language processing on realworld questions. We hope that this dataset will lead to wider research interest in this important health domain.

M3 - Other contribution

VL - 193

ER -