Term-Based Approach for Linking Digital News Stories

Abstract

The World Wide Web has become a platform for news publication in the past few years. Many television channels, magazines and newspapers have started publishing digital versions of the news stories online. It is observed that recommendation systems can automatically process lengthy articles and identify similar articles to readers based on a predefined criteria i.e. collaborative filtering, content-based filtering approach. The paper presents a content-based similarity measure for linking digital news stories published in various newspapers during the preservation process. The study compares similarity of news articles based on human judgment with a similarity value computed automatically using common ratio measure for stories. The results are generalized by defining a threshold value based on multiple experimental results using the proposed approach.