Estimating and analyzing the popularity of an entity is an important task for professionals in several areas, including the music industry and social media. As the FuturePulse team points out in our project’s newest paper ‘GAP: Geometric Aggregation of Popularity Metrics’, the abundance of online data should enhance peoples’ insights regarding the collective human behavior.

Nevertheless, effectively modeling popularity and integrating diverse data sources, are still considered to be very challenging – and yet there is actually no optimal approach on how to tackle this challenge.

Here is where our very own Christos Koutlis, Manos Schinas, Symeon Papadopoulos and Ioannis Kompatsiaris from the Information Technologies Institute of the Centre of Research and Technology Hellas come, proposing the Geometric Aggregation of Popularity metrics (GAP), ‘a non-linear method for popularity metric aggregation, based on geometrical shapes derived from the individual metrics’ values’.

In specific, the team focuses on the estimation of music artist popularity by aggregating web-based artist popularity metrics, derived from social media and streaming platforms. As the authors point out, this was the first attempt to aggregate multiple popularity sources in the academic literature, related to music information retrieval. And most importantly, it has actually yielded satisfactory results on summarizing an artist’s popularity picture.

As the FuturePulse team analyses: ‘the algorithm used geometrical shapes formatted by the individual metrics’ values of each entity, and it was found to outperform the most natural choice for metric aggregation, being a simple average, with respect to several measures of similarity between the computed metrics and reference data’. And what is more extraordinary? The proposed aggregation method was robust even when the under study artist was popular only in some of the monitored popularity metrics!

In addition, the authors mention that even though the most natural choice for metric aggregation would be a linear model, their approach leads to ‘stronger rank correlation and non-linear correlation scores compared to linear aggregation schemes’. More precisely, the FuturePulse team’s approach outperforms the simple average method in five out of seven evaluation measures, described in the section ‘3.3 Evaluation’ of the paper.

Last but not least, it is pinpointed that this methodology could be furthermore extended for use in several other areas, such as cinema and football in which actors and players will serve as entities and their social media accounts and other related factors like tickers sold, as metrics. As our FuturePulse partners explain ‘Future work will include the evaluation of all metric aggregation methods on other tasks, such as the prediction of individual metrics’ future values’.

 

You can read the paper ‘GAP: Geometric Aggregation of Popularity Metrics’ here.