Abstract— With an increasing number of documents for drug information, automatic classification of documents is an important task for organizing these documents into appropriated classes. Only a few research works on text classification of drug documents, were contributed. In this task, monographs of drug can be categorized to one class or multiple classes by their indications. A two-step text classification for multi-label documents is proposed. A set of drug monographs is drawn from RxList and DailyMed websites and used for evaluating the proposed method. In the first step, documents are classified as single-label or multi-label documents. The result from the first step is applied on the second step of classification based on their therapeutic classes. From the experimental results, the two-step classification is an efficient method to classify a document into one or multiple classes.
Index Terms— Text categorization, text classification, multi-label classification, drug monograph classification.
The authors are with the Faculty of Pharmacy, Silpakorn University, Nakorn Pathom, 73000 Thailand (e-mail: verayuth@su.ac.th, verayuths@hotmail.com, chanisara@su.ac.th).
[PDF]
Cite: Verayuth Lertnattee and Chanisara Lueviphan, " Two-Step Text Classification for Multi-Label Drug Monographs," International Journal of Innovation, Management and Technology vol. 4, no. 2, pp. 272-276, 2013.