Automatic summarization of Turkish documents using non-negative matrix factorization

Automatic document summarization is a process, where a computer summarizes a document. This paper presents the performance analysis of an automatic Turkish document summarization system that applies Non-negative matrix factorization based summarization algorithm with different preprocessing methods. The preprocessing method called “Consecutive Words Detection” is an innovative approach that uses Turkish Wikipedia links to represent related consecutive words as a single term and the result of the evaluation process is promising for document summarization in Turkish.