The Design Of Railway Public Opinion Monitor System Based On Hadoop

Railway is an important national transportation infrastructure. It also provides a popular means of transportation for the public and acts as an artery of national economy. In 2013, the China Railway Corporation enforces energetic reform through launching a series of measures for customers, such as Internet ticket booking, telephone ticket booking, real name system, e-commerce on freight which also bring a series of criticism and different opinions in the society. In order to better understand the social conditions and public opinion and improve railway work and service quality, it is of great significance to build a public opinion monitoring system with public opinion monitoring, analysis, judgment, coordination and disposal, comprehensive evaluation and report functions.The thesis analysis and summary of the development and the present situation of both domestic and foreign public opinion monitoring and points out the problems existing in current railway public opinion monitoring system. According to the current demand of railway public opinion detection, the collected data and functions were analyzed. Public opinion data has two main characteristics which are massive and unstructured. Due to the limitations of technology, traditional technology can not be a good response. Hadoop is a open source big data platform, which was specifically designed to deal with massive data. HDFS, Hbase and MapReduce are core modules of Hadoop. The thesis then makes a design of high performance, safe and reliable public opinion monitoring processing platform based on these modules. It can meet the actual needs of the railway public opinion monitor.