Dermatological diseases are among the most common medical conditions, affecting nearly 900 million people worldwide. Despite their prevalence, they remain among the most neglected diseases. Dermatological diseases can largely impact patients’ quality of life and pose serious repercussions. Therefore, accurate diagnosis and treatment by healthcare professional, such as a dermatologists, are essential. However, the ambiguous use of terminology in the textual descriptions of dermatological conditions and their associated signs/symptoms, poses a challenge for healthcare professionals in correctly identifying the underlying skin diseases. To address this challenge, we propose a corpus-based approach to develop a lexicon of dermatological diseases. The proposed framework utilizes a seed word list extracted from PubMed abstracts. This lexicon is further enhanced with UMLS knowledge sources (UMLS Metathesaurus and SPECIALIST Lexicon) to incorporate synonymous terms and concept variants, thereby enhancing the quality and coverage of the generated lexicon. Additionally, the lexicon is evaluated on drug reviews from two benchmark datasets (Druglib.com and WebMD.com), by classifying the reviews into dermatological and non-dermatological disease categories. The specialized lexicon enables dermatology researchers and professionals to uncover critical information about dermatological diseases, ultimately improving the delivery of dermatological care.