Mining Useful News Information Based on Reader Feedback for Building News Communities

Abstract

Due to the rapid development of information and communication technologies (ICTs), the Internet has become one of the most important communication media for journalism. Models of reader reception of information have changed with websites providing convenient and interactive user interfaces for communicating information instantly. However, this evolution has generated an unprecedented challenge for traditional communication media such as print newspapers. Online news sites need to create unique content to attract readers, and need to develop engaging community management services with Web 2.0 interactive mechanisms to compete with other websites and retain user attention. This work presents a novel online news platform that facilitates the construction of a University Press reader community. This platform can automatically analyze the reader community dataset of University newspapers, including an opinion deviation indicator, popularity indicator, and topicality indicator for each news story. Based on fuzzy inference, the proposed online news platform can select targeted news stories using these three indicators to identify top news stories that promote debate and interactivity within a reader community and promote communication efficiency and reader engagement. Experimental results reveal that the proposed interactive mechanisms satisfy the needs of most readers and correctly display top news stories that readers find interesting. Additionally, the proposed online news platform can assist journalists in understanding reader needs while promoting online social interaction.