Feedback Matters! Predicting the Appreciation of Online Articles A Data-Driven Approach - Machine Learning and Knowledge Extraction Access content directly
Conference Papers Year : 2018

Feedback Matters! Predicting the Appreciation of Online Articles A Data-Driven Approach

Abstract

The current era of advanced computational mobile systems, continuous connectivity and multi-variate data has led to the deployment of rich information settings that generate constant and close to real-time feedback. Journalists and authors of articles in the area of Data Journalism have only recently acknowledged the influence that the audience reactions and opinions can bring to effective writing, so to be widely appreciated. Such feedback may be obtained using specific metrics that describe the user behavior during the interaction process like shares, comments, likes, claps, recommendations, or even with the use of specialized mechanisms like mood meters that display certain emotions of readers they experience while reading a story. However, which characteristics can reveal an article’s character or type in relation to the collected data and the audience reflection to the benefit of the author? In this paper, we investigate the relationships between the characteristics of an article like structure, style of speech, sentiment, author’s popularity, and its success (number of claps) by employing natural language processing techniques. We highlight the emotions and polarity communicated by an article liable to increase the prediction regarding its acceptability by the audience.
Fichier principal
Vignette du fichier
472936_1_En_10_Chapter.pdf (369.19 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02060049 , version 1 (07-03-2019)

Licence

Attribution

Identifiers

Cite

Catherine Sotirakou, Panagiotis Germanakos, Andreas Holzinger, Constantinos Mourlas. Feedback Matters! Predicting the Appreciation of Online Articles A Data-Driven Approach. 2nd International Cross-Domain Conference for Machine Learning and Knowledge Extraction (CD-MAKE), Aug 2018, Hamburg, Germany. pp.147-159, ⟨10.1007/978-3-319-99740-7_10⟩. ⟨hal-02060049⟩
56 View
97 Download

Altmetric

Share

Gmail Facebook X LinkedIn More