One of the biggest social networks is Twitter for providing message posting and use direct message services via “tweets”.366 million currently active users on Twitter, Is the second largest space to share common news or message post. These features are also used by spammers on Twitter, spammers are not new on Twitter. spammers must be detected in improving the quality of Twitter message services. In this spammer detected by using metadata, content, interaction, and community-based features methods. Tweet meta information extracted and analyzed based on user-id, tweets,tweet-time and tweet-type. Content Features are extracted based on user posting content with URL, Mention-tags and hash-tags. Interaction and community-based features are analyzed by following and follower information. The proposed approach to introducing network-based features for spammer ratio detection by using unique IP address based user classification. Spammers can be detected by analyzing their tweets based on the extended hybrid approach by using the random forest, decision tree, and Bayesian network on the Twitter dataset that has benign users and spammers.