Abstract
Large number of Internet users in the Arab World uses the Internet as their main source of information. There are many news publishers and agencies that compete to gain the largest percentage of Arab visitors to their Web sites. Also these agencies received a continuous stream of news articles each day, and therefore there is a need for an automated system to identify new news items to inform the employee in these agencies about it, while this system has to neglect old news articles. This study demonstrates a novelty detection system, which used to identify the topic of each Arabic news item, besides identifying the novelty of each Arabic news item.