Design of a Support Vector Machine Aided Real Time Health Information Text Extraction System

Abstract

The exponential growth of available information sources has greatly affected the access to useable health information. As a consequence, medically biased information has become difficult to use for decision making. In this paper, we consider these consequences and present an enhanced method for accessing health information in real time. The approach involves the use of the vapnik Support Vector Machine process for text classification. The proposed method was frameworked on php/mysql for web user. Experimental setup shows that the method outperforms the baseline in the Precision, Recall and F1 measures. An extension using the Gaussian kernel is recommended in the paper