E. Query Expansion
Research done in [17] focusing on a new query
expansion method which combined the ontology-based
collaborative filtering and neural networks to increase the
performance of expansion method. Collaborative filtering is
a method that assesses similarities among the previous users
then recommends the documents that are accessed by the
users to the current users [18]. Example of collaborative
filtering is in Fig. 1. But the limitation of the collaborative
filtering is that it cannot find any connection between users
and their similar users when little information was given by
the user at the beginning of the search. Researchers in [17]
proposed the use of ontology-based collaborative filtering to
analyze semantic relationships between similar users where
these external ontologies can provide some personal
information about all the users and their interests.