*Article* **Topic Modeling for Analyzing Patients' Perceptions and Concerns of Hearing Loss on Social Q&A Sites: Incorporating Patients' Perspective**

**Junghwa Bahng 1,\* and Chang Heon Lee 2,\***


Received: 27 June 2020; Accepted: 24 August 2020; Published: 27 August 2020

**Abstract:** Hearing loss is the most common human sensory deficit, a ffecting normal communication. Recently, patients with hearing loss or at risk of hearing loss are increasingly turning to the online health community for health information and support. Information on health-related topics exchanged on the Internet is a useful resource to examine patients' informational needs. The ability to understand the patients' perspectives on hearing loss is critical for health professionals to develop a patient-centered intervention. In this paper, we apply Latent Dirichlet Allocation (LDA) on electronic patient-authored questions on social question-and-answer (Q&A) sites to identify patients' perceptions, concerns, and needs on hearing loss. Our results reveal 21 topics, which are both representative and meaningful, and mostly correspond to sub-fields established in hearing science research. The latent topics are classified into five themes, which include "sudden hearing loss", "tinnitus", "noise-induced hearing loss", "hearing aids", "dizziness", "curiosity about hearing loss", "otitis media" and "complications of disease". Our topic analysis of patients' questions on the topic of hearing loss allows achieving a thorough understanding of patients' perspectives, thereby leading to better development of the patient-centered intervention.

**Keywords:** hearing loss; latent topic; LDA; topic modeling; social Q&A
