Spam Identification in Cloud Computing Based on Text Filtering System

With the rapid increase of spam on Internet and diversification its forms, how to quickly effectively identify a large number has become an urgent topic. Cloud computing obvious advantages in storage processing, so it can calculate amount mail data. Due uncertainty life cycle spam, feedback re-judgment is added anti-spam system, text filtering system based active learning with four stages training, filtering, feedback, re-filtering implemented. Compared original improve keywords. In order reduce misjudgment rate ordinary accuracy judgment, suggested use weighted decision-making email header information implement effective auxiliary classification. For emails lacking content, method title weighting feasible effective, which identification relatively little content. Because cloud far more advanced than traditional algorithm, development solve infinite spam. Therefore, this paper makes in-depth study by summarizing analyzing current technologies.