Predicting Key Events in the Popularity Evolution of Online Information

Abstract

The popularity of online information generally experiences a rising and falling evolution. This paper considers the “burst”, “peak”, and “fade” key events together as a representative summary of popularity evolution. We propose a novel prediction task—predicting when popularity undergoes these key events. It is of great importance to know when these three key events occur, because doing so helps recommendation systems, online marketing, and containment of rumors. However, it is very challenging to solve this new prediction task due to two issues. First, popularity evolution has high variation and can follow various patterns, so how can we identify “burst”, “peak”, and “fade” in different patterns of popularity evolution? Second, these events usually occur in a very short time, so how can we accurately yet promptly predict them? In this paper we address these two issues. To handle the first one, we use a simple moving average to smooth variation, and then a universal method is presented for different patterns to identify the key events in popularity evolution. To deal with the second one, we extract different types of features that may have an impact on the key events, and then a correlation analysis is conducted in the feature selection step to remove irrelevant and redundant features. The remaining features are used to train a machine learning model. The feature selection step improves prediction accuracy, and in order to emphasize prediction promptness, we design a new evaluation metric which considers both accuracy and promptness to evaluate our prediction task. Experimental and comparative results show the superiority of our prediction solution.

A Prediction Method of Peak Time Popularity Based on Twitter Hashtags

“…This work is a follow-up of our previous study that provided a conventional machine learning method (Support Vector Regression, SVR) [39]. We here extend it by conducting some empirical studies and providing a multi-modal deep learning method with better performance.…”

“…What we need to say here is that this paper takes absolute error (the difference between the ground truth value and the predicted value) as the metric for Q3. All predictions are triggered before popularity reaches 40, as explained in [39].…”

“…As we did in previous work [39], this paper feeds SVR handcrafting features, such as network topological features (the average node degree, the maximum node degree, the global clustering coe?cient, etc.) and hashtag string features (the number of words and hashtag string lengths).…”

When a luxury brand bursts: Modelling the social media viral effects of negative stereotypes adoption leading to brand hate

“…Specifically, popularity is defined as a piece of online information receiving a certain level of attention in social media, which can be measured through the number users who actually viewed the message, or the number of comments achieved, likes, retweet, etc. The information shared online exhibits different popularity evolution, since it is impacted by many factors such as celebrity involvement, a certain event occurring in the real world, user interests, and so on ( Hu, Hu, Fu, Fang, & Xu, 2017 ). Accordingly, recent studies tried to understand why certain information become more popular than others, by considering the richness in informational content ( Araujo, Nijens, & Vliegenthart, 2015 ), vividness and interactivity ( de Vries et al, 2012 ), emotional involvement ( Pantano, Giglio, & Dennis, 2019 ), narrative styles ( Aleti et al, 2019 ), originality and uniqueness ( Casalo et al, 2020 ), and presence of certain visual elements ( Villarroel Ordenes et al, 2019 ).…”

“…The diffusion of some pieces of online information also shows analogies with the diffusion (transmission) of infectious diseases (contagion) ( Rapp et al, 2013 ). In the both cases, the contagion starts with a number of entities/users who diffuse the virus/message ( Hu et al, 2017 ). Subsequently, the number of users liking/commenting/replying to the message represents the number of users “infected” by the message, which becomes viral/highly popular as in the unlucky case of infectious diseases ( Kiss and Bichler, 2008 , Rapp et al, 2013 ).…”