Recommendation system for web article based on association rules and topic modelling

Abstract

The World Wide Web is now the primary source for information discovery. A user visits websites that provide information and browse on the particular information in accordance   with their   topic interest.   Through  the  navigational process,  visitors  often  had  to  jump  over  the  menu  to  find  the right  content.  Recommendation system can help the visitors to find the right content immediately.  In this study, we propose a two-level recommendation system, based on association rule and topic similarity.  We generate association rule by applying Apriori algorithm.   The  dataset  for  association  rule  mining  is a  session of  topics  that  made  by  combining  the  result of  sessionization and  topic  modeling.  On  the  other   hand,   the  topic  similarity made  by  comparing   the  topic  proportion of  web  article.  This topic proportion inferred from the Latent Dirichlet Allocation (LDA). The results show that in our dataset there are not many interesting   topic relations in one session.  This  result  can  be resolved,  by  utilizing  the  second  level  of  recommendation  by looking into the article  that  has the similar  topic.