Semantic Tagging-Based Document Retrieval Using Non-Negative Matrix Factorization

Abstract

Many document retrieval methods focusing on unstructured text to deliver more meaningful information on the user. Tag-based document retrieval aims to address a challenge to searching relevant text-documents given a set of tags. Tag-based approaches received a wide attention as a possible solution to the big-content related IR, showing a high performance through a combination of its effectiveness and efficiency. This paper use word sense disambiguation with non-negative matrix factorization to generate topic model based semantic.