Anti spam algorithm is to identify

Anti spam algorithm is to identify and process some spam ads intelligently by big data algorithm. Figure 3-6 shows the overall framework of Meiyou's anti garbage algorithm using big data, which consists of two parts. Above the dotted line is the training process of anti spam algorithm, which is initially based on NLP natural language processing. First, the text data (junk post and normal post) is segmented, which needs to be updated regularly, then the post is processed and selected, and the extracted features are sent to the classifier model for training, including Bayesian classification and logical regression classification And output the result of classification model through training. These trainings were initially carried out in their own computer rooms. Later, with the increase of data volume, some model trainings have been transferred to Alibaba cloud.