Topic Trend in Aceh Online Newspaper Based on Significant Term
Arie Budiansyah, Viska Mutiawani · 2014
Abstract. Online newspaper is very common these days and provides information faster than hardcopy newspaper. Online newspaper can be accessed very easily and mostly free. Lots of articles are published every day in online newspaper so we can get lots of information from it. Text and web mining is a way to gain information from online newspaper. This research is trying to find topic trend in Aceh online newspaper based on significant term. This research uses articles from “Serambi Indonesia”, one of newspaper that has online version, as the online newspaper data source. This research has mined 28.071 articles in one year from Serambi Indonesia online newspaper. The research approach has two steps: (i) Data preprocessing; (ii) Significant term extraction and term history generation. The first step uses agent to download the articles automatically and extracts the content of those articles. The second step uses an existing external memory approach to extract significant terms while computing term history simultaneously. Significant term is a series of words that are significant enough to represent one event, action or concept. Term history of one significant term is the term frequency over consecutive time periods. This research has found out 1.402.634 significant terms and stored them in database. The terms can be searched and viewed as graphics divided by months as period division. So by knowing the significant term, this research provides topic trend in Aceh online newspaper in monthly time. Keywords: online newspaper, significant term, topic trend, text and web mining