Klasifikasi Teks Pengaduan Masyarakat Dengan Menggunakan Algoritma Neural Network

Abstract

The development of the Internet led to a flood of digital information. Various information can be obtained easily just by clicking or pressing 'enter'. Dissemination of information in the form of digital documents has experienced unprecedented growth. One of the web intended for the general public in the province of Central Java, web named 'Lapor Gub!'. This site was made with the intention to accommodate the aspirations, concerns, or complaints against public services and also dissatisfaction with the performance of local government and the provincial government of Central Java. An increasing number of documents in text format significantly recently made the process of grouping documents (document classification) becomes important. By using the method of text classification, then the document is an overwhelming number of these are organized in such a way so as to facilitate and accelerate the search needed information. Experiments in this study aimed to classify the Indonesian language text documents using Neural Network algorithm. The test is done by using a sample of text documents taken from a web-based electronic mass media entitled 'Lapor Gub!'. The experimental results show that the neural network method effectively used to classify texts public complaints. It is seen from the experimental results, namely the use of Neural Network algorithm on the classification process produces high accuracy in the amount of 43.00% with a period of 03 hours 45 minutes 14 seconds of the Indonesian language text documents classify text public complaints