Research Article
A Prerecognition Model for Hot Topic Discovery Based on Microblogging Data
| when after each crawling : | | if new then | | send keyword dictionary, stop words, filter words to | | processing model | | split to keywords, post id set, | | and filter for processing | | put to topic model, | | use LDA, PAM to generate | | cluster with model | | if then | | update exist do | | | | else | | | | | | if then | | | | calculate , , , | | if then | | label as | | if then | | label as | | if then | | label as | | if then | | label as | | if then | | label as | | if then | | label as | | send topics at for ranking | | select hot topic when |
|
