Mining the Main Health Trend of the General Public based on Opinion Mining of Korean Blogsphere

Abstract

These days, social media usually becomes a reasonable standard for understanding the public's thought. Especially, people increasingly use internet media and SNS (twitter, facebook, blog, and etc.), to share opinions, news, advice, interests, moods, concerns, critics, facts, rumors, and everything. Therefore, public health research has been started a big change. Traditional public health study has depended on only regular clinical reports by health professionals. It is limited to practical use and general public has much difficulty to understand health information, even if it's his/her own information. Nowadays, over one billion people publish their ideas about many topics, including health conditions minute by minute. SNS provides researchers the freshest source of public health conditions on a global scale. Much of that data is public and available for mining. So this article pursues making an application of opinion mining for detecting the public's trend and finding valuable opinion among the massive information. The core of this research is analyzing the adjective of opinions. Our assumption is that many adjective expressions implicate deep and sincere meaning of its author. It is applicable for both low value postings filtering and tracking high value postings simultaneously. This approach is a simple and feasible criteria. The opinion mining process includes Korean morpheme analysis, opinion extraction, opinion tagging, positive / negative score evaluation. Our research's aim is to analyze Korean blog postings.