Building weather season pattern from Aceh local news article (Text period analysis via significant term)

摘要

Knowledge about climate change is very important in human life. The information gained can be used to learn about time period of weather season from January to December. Indonesia is a tropical country with two weather seasons, they are summer and rainy weather. Weather information and forecasting are usually reported by Indonesian Agency for Meteorology, Climatology and Geophysics, BMKG (Badan Meteorologi Klimatologi dan Geofisika), every day on television. The method used by BMKG i.e. using weather sensor equipment placed in several locations to grab the weather data. This study also grabs the weather data by mining a text in local electronic newspaper. The idea of this research came from reading a few local newspaper articles that tells about impact of weather event such as droughts, floods, crop failure, somewhere then we know each article has timestamp such as the date, month, and year. Therefore by looking "the word that tells about impact of weather event" in electronic newspaper article, from January to December, then we know weather conditions. Those word called weather significant terms. To gain this information, we classified several word which can represent summer and rainy weather condition. These words were collected from three stages of text mining process: (i) Data preprocessing; (ii) Sorting suffixes and saving into file; (iii) Extracting weather significant term candidate along with timestamp. Weather significant term and their timestamps will be stored in a database. Later, an application will query the database to find information of weather season words (weather significant terms). The pattern will use month as the interval period. By entering some weather significant terms related to summer and rainy weather season into application, the application will produce graph explaining summer and rainy weather season pattern period in Aceh province.