Generating and Evaluating Text Summarisations using Text-representing Centroids(TRC)

Abstract

Short abstracts and text summarisation are gaining increasing the importance of tools for filtering out the most relevant articles due to the increased documents and information. The following article uses TRCs to support text summaries’ generation and evaluation. A new method of document summarisation has been introduced based on text-representing centroids (TRC). TRC-based similarity measure delivers good similarity estimations for both methods, which are in the excellent range of 75 percent on average.