Automatic Estimation of News Values Reflecting Importance and Closeness of News Events

Abstract

This paper addresses a problem of automatic estimation of three journalistic news values, more specifically frequency, threshold and proximity, by applying various text mining methods. Although theoretical frameworks already exist in social sciences that identify if an event is newsworthy, these manual techniques require enormous amount of time and domain knowledge. Thus, we illustrate how text mining can assist journalistic work by finding news values of different international publishers across the world. Our experiments both on a collection of news articles from different publishers about Apple's launch of new iPhone 6 and Apple Watch and on a wider collection of documents confirm that some journalists still follow some of the well-known journalistic values. Furthermore, we acknowledge that news values are often orthodox and outdated, and no longer apply to all publishers. We also outline possible future implications of our approach to work on interaction between text mining and journalistic domains. Povzetek: Članek obravnava problem avtomatske ocene novičarskih vrednosti, natančneje: pogostosti, prag pomembnosti ter bližine (oz. relevance), z uporabo različnih metod avtomatske analize teksta. Čeprav v družboslovju obstaja več teoretičnih ogrodij, ki določajo ali je nek dogodek vreden poročanja, te temeljijo na »ročnem« delu in zahtevajo veliko časa ter globoko poznavanje domene. V članku nakažemo, kako lahko avtomatska analiza teksta pomaga pri novinarskem delu z uvidom v novičarske vrednosti na globalnem nivoju in med velikimi, mednarodnimi mediji. Rezultati naših poskusov na zbirki člankov iz različnih virov (spletnih časopisov) o splovitvi izdelkov iPhone 6 in Apple Watch podjetja Apple potrjujejo, da novinarji še sledijo nekaterim uveljavljenim novičarskim vrednostim. Ob tem ugotavljamo, da so nekatere novičarske vrednosti preveč ortodoksne in zastarele ter ne veljajo več za vse novičarske vire. Na koncu orišemo načrtovano nadaljnje delo in možne implikacije uporabe avtomatskih metod analize teksta na novinarsko delo.

Adapting the influences of publishers to perform news event detection

“…One of the key contributions of this article is a module for finding trigger words in news. Belyaeva et al [7] suggested that the number of publishers can be used as an indicator when evaluating the importance of news. Therefore, this study adopts three methods to extract trigger words and proposes a method named the 'pyramid relation algorithm' to obtain event relation words.…”

“…Accordingly, this study aims to extract the features and words from news datasets that are considered valuable by the news media. However, there is much related research that uses text analysis techniques to deal with the vast amounts of news for topic classification or clustering [1,4,[7][8][9][10][11]. Classification is the process of categorising data with the help of class labels.…”

“…Notably, only valuable news that attracts public concerns and produces resonance is reported by journalists and officially released. The measurement of news value comes from the subjective judgement of journalists and media organisations [7]. Accordingly, this study aims to extract the features and words from news datasets that are considered valuable by the news media.…”

Analysis and Design of Computational News Angles

“…To the best of our knowledge, news angles have not yet been investigated in the research literature from a computational perspective, with the notable exception of the aforementioned work on automatic news classification with respect to news values [5], [6]. In the media industry, the IPTC's NewsML G2, EventsML G2, and rNews formats [21] and the BBC ontologies (https://www.bbc.co.uk/ontologies) offer standards for representing and exchanging news-related information; however, none of them considers news angles.…”

“…Finding good angles on potentially newsworthy events and situations is therefore an important journalistic task, which needs to be better supported by journalistic software tools and platforms. However, although news angles and related concepts are well covered in the journalism literature on a general level, there has been so far only limited computational treatment of these notions, which has primarily focused on the automatic identification of the related notion of news values 1 in journalistic content [5], [6]. In particular, a comprehensive framework for representing and reasoning with news angles is still missing.…”