Predicting Clickbait Strength in Online Social Media

Proceedings of the 28th International Conference on Computational Linguistics, pages 4835–4846 Barcelona, Spain (Online), December 8-13, 2020 4835 Predicting Clickbait Strength in Online Social Media Vijayasaradhi Indurthi∗, Bakhtiyar Syed, Manish Gupta†, Vasudeva Varma IIIT Hyderabad, India {vijaya.saradhi,syed.b}@research.iiit.ac.in {manish.gupta,vv}@iiit.ac.in Abstract Hoping for a large number of clicks and potentially high social shares, journalists of various news media outlets publish sensationalist headlines on social media. These headlines lure the readers to click on them and satisfy the curiosity gap in their mind. Low quality material pointed to by clickbaits leads to time wastage and annoyance for users. Even for enterprises publishing clickbaits, it hurts more than it helps as it erodes user trust, attracts wrong visitors, and produces negative signals for ranking algorithms. Hence, identifying and flagging clickbait titles is very essential. Previous work on clickbaits has majorly focused on binary classification of clickbait titles. However not all clickbaits are equally clickbaity. It is not only essential to identify a click- bait, but also to identify the intensity of the clickbait based on the strength of the clickbait. In this work, we model clickbait strength prediction as a regression problem. While previous methods have relied on traditional machine learning or vanilla recurrent neural networks, we rigorously investigate the use of transformers for clickbait strength prediction. On a benchmark dataset with ∼39K posts, our methods outperform all the existing methods in the Clickbait Challenge1. 1 INTRODUCTION Clickbait refers to those sensational, provocative or controversial posts which appear to be informative and objective, but are designed to entice its readers into clicking the link accompanying with the post. Fig. 1 shows popular clickbait types with examples. According to a survey of 53 Stanford students, 96.2 percent of Stanford students encounter clickbait articles on the Internet at least once per day2. Rony et al. (2017) estimate that 19.46% of headlines were “clickbait” in 2014; 23.73% in 2015; and 25.27% in 2016. Beyond the increased prevalence, clickbait is also a challenge across multiple modes of data, text, images and even videos3. The economic model of the contemporary online news industry (Dvorkin, 2015) incentivizes more content views. A report by the Columbia Journalism Review highlighted the case of online magazine Slant, which pays writers $100 per month, plus $5 for every 500 clicks on their stories. This clearly motivates journalists to write catchy and suspenseful headlines. Table 1 lists some of the popular social media outlets publishing clickbait content and their followers. The numbers are indicative of how much people are easily falling to the bait. Cognitively, human minds have a tendency to satisfy and bridge their curiosity gap by clicking on the link. Marketing companies have been using clickbaits to attract and engage more number of users resulting in getting more page views. Websites need more page views to promote their content or to create more opportunities to show advertisements which increase their revenue. Moreoever, it is a well This work is licensed under a Creative Commons Attribution 4.0 International License. License details: http://creativecommons.org/licenses/by/4.0/. ∗The author …

Content truncated to 500 words