TY - JOUR
T1 - A deep learning architecture for psychometric natural language processing
AU - Ahmad, Faizan
AU - Abbasi, Ahmed
AU - Li, Jingjing
AU - Dobolyi, David G.
AU - Netemeyer, Richard G.
AU - Clifford, Gari D.
AU - Chen, Hsinchun
N1 - Funding Information:
The authors wish to thank the U.S. National Science Foundation for their support under grants NSF IIS-1816504, IIS-1553109, BDS-1636933, CCF-1629450, and Microsoft Research for its support through CRM:0740129. Authors’ addresses: F. Ahmad, 14 University Circle #4, Charlottesville, VA 22903; email: fa7pdn@virginia.edu; A. Abbasi, J. Li, D. G. Dobolyi, and R. G. Netemeyer, Rouss Hall and Robertson Hall (McIntire School of Commerce), 125 Ruppel Dr, Charlottesville, VA 22903; emails: {abbasi, jl9rf, dd2es}@comm.virginia.edu, rgn3p@virginia.edu; G. D. Clifford, Department of Biomedical Engineering, 313 Ferst Drive, Room 2127 Atlanta, GA 30332; email: gari.clifford@bme.gatech.edu; H. Chen, McClelland Hall 430X 1130 E. Helen St. Tucson, Arizona 85721-0108; email: hchen@eller.arizona.edu.
Publisher Copyright:
© 2020 Copyright held by the owner/author(s).
PY - 2020/2/5
Y1 - 2020/2/5
N2 - Psychometric measures reflecting people's knowledge, ability, attitudes, and personality traits are critical for many real-world applications, such as e-commerce, health care, and cybersecurity. However, traditional methods cannot collect and measure rich psychometric dimensions in a timely and unobtrusive manner. Consequently, despite their importance, psychometric dimensions have received limited attention from the natural language processing and information retrieval communities. In this article, we propose a deep learning architecture, PyNDA, to extract psychometric dimensions from user-generated texts. PyNDA contains a novel representation embedding, a demographic embedding, a structural equation model (SEM) encoder, and a multitask learning mechanism designed to work in unison to address the unique challenges associated with extracting rich, sophisticated, and user-centric psychometric dimensions. Our experiments on three real-world datasets encompassing 11 psychometric dimensions, including trust, anxiety, and literacy, show that PyNDA markedly outperforms traditional feature-based classifiers as well as the state-of-the-art deep learning architectures. Ablation analysis reveals that each component of PyNDA significantly contributes to its overall performance. Collectively, the results demonstrate the efficacy of the proposed architecture for facilitating rich psychometric analysis. Our results have important implications for user-centric information extraction and retrieval systems looking to measure and incorporate psychometric dimensions.
AB - Psychometric measures reflecting people's knowledge, ability, attitudes, and personality traits are critical for many real-world applications, such as e-commerce, health care, and cybersecurity. However, traditional methods cannot collect and measure rich psychometric dimensions in a timely and unobtrusive manner. Consequently, despite their importance, psychometric dimensions have received limited attention from the natural language processing and information retrieval communities. In this article, we propose a deep learning architecture, PyNDA, to extract psychometric dimensions from user-generated texts. PyNDA contains a novel representation embedding, a demographic embedding, a structural equation model (SEM) encoder, and a multitask learning mechanism designed to work in unison to address the unique challenges associated with extracting rich, sophisticated, and user-centric psychometric dimensions. Our experiments on three real-world datasets encompassing 11 psychometric dimensions, including trust, anxiety, and literacy, show that PyNDA markedly outperforms traditional feature-based classifiers as well as the state-of-the-art deep learning architectures. Ablation analysis reveals that each component of PyNDA significantly contributes to its overall performance. Collectively, the results demonstrate the efficacy of the proposed architecture for facilitating rich psychometric analysis. Our results have important implications for user-centric information extraction and retrieval systems looking to measure and incorporate psychometric dimensions.
KW - Deep learning
KW - Natural language processing
KW - Psychometric measures
KW - Text classification
UR - http://www.scopus.com/inward/record.url?scp=85079433738&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85079433738&partnerID=8YFLogxK
U2 - 10.1145/3365211
DO - 10.1145/3365211
M3 - Article
AN - SCOPUS:85079433738
VL - 38
JO - ACM Transactions on Office Information Systems
JF - ACM Transactions on Office Information Systems
SN - 1046-8188
IS - 1
M1 - 3365211
ER -