Predicting long-term user engagement

Interest as a Proxy of Engagement in News Reading: Spectral and Entropy Analyses of EEG Activity Patterns

 I. Arapakis, M. Barreda-Àngeles, and A. Pereda-Baños

Abstract - Objective measurements of engagement are increasingly sought after by both the media industry and scholar communities to explain what drives people to consume audiovisual contents. However, engagement is a complex construct that, at the psychological level, has been mainly operationalised through indicators of attentional and emotional processes, often overlooking motivational factors. We claim that in the context of news consumption, motivation, operationalised as intrinsic interest for consuming a given content, needs to be factored in together with attentional and emotional processes. The present work provides an objective metric for motivation based on electroencephalographic (EEG) registration of users’ neural activity, while they read sets of news pre-classified in terms of their potential interest. We focus on a metric that has been used as an indicator of the degree to which an item or event induces the motivation to approach or escape, the so called frontal alpha asymmetry (FAA). Moreover, in addition to the traditional approach to the analysis of EEG signals, we also introduce a more novel technique based on estimating the entropy of the signals. Results confirm that FAA is indeed a good proxy for objective monitoring of interest in media contents and that entropy analysis, although its interpretation in terms of information processing warrants further investigation, is also sensitive to the manipulation of interest, providing results that complement traditional power spectrum analysis.

 PDF
 User Engagement; News Consumption; EEG; Spectral Analysis; Entropy Analysis; Predictive Modelling

(a)
(b)

Example (participant 50) of the relative activation (for different frequency bands in Hz) of the frontal channel (FAA) across the experimental conditions.

Variation of user engagement measures across conditions (interaction with news genres).

© Ioannis Arapakis 2016 with help from Bootstrap