Identifying Journalistic Angles

Abstract

Journalism is facing significant financial challenges, with declining advertisement and loss of revenues while the increased pressure and competition from free online distribution channels pushes on. Despite this, digital consumption is in high demand and the need for quality journalism based on trusted sources and editorial work is also increasing. Examples of this need can be highlighted by events such as the COVID-19 pandemic and BREXIT.

This thesis aims to research how computational journalism and machine learning techniques can aid in journalistic news production. This is explored by analyzing a human annotated dataset on news articles and news angles. A framework for classifying news angles is proposed and tested out as a base for survey experiments.