How popular is a topic on social media? A multi-criteria decision-making framework based on user engagement

Abstract

Purpose

Social media is widely used to capture citizens' opinions and topics deemed important. The importance or interest social media users attribute to a topic is traditionally measured by tweet frequency. This approach is practical but overlooks other user engagement tools such as retweets, likes, quotes, and replies. As a result, it may lead to a misinterpretation of social media signals. This paper aims to propose a method that considers all user engagement indicators and ranks the topics based on the interest attributed by social media users.

Design/methodology/approach

A multi-criteria decision-making framework was proposed, which calculates the relative importance of user engagement tools using objective (information entropy) and subjective (Bayesian Best-Worst Method) methods. The results of the two methods are aggregated with a combinative method. Then, topics are ranked based on their user engagement levels using Multi-Objective Optimization by Ratio Analysis.

Findings

The proposed approach was used to determine citizens' priorities in transport policy, and the findings are compared with those obtained solely based on tweet frequency. The results revealed that the proposed multi-criteria decision-making framework generated more comprehensive and robust results.

Practical implications

The proposed method provides a systematic way to interpret social media signals and guide institutions in making better policies, hence ensuring that the demands of users/society are properly addressed.

Originality/value

This study presents a systematic method to prioritize user preferences in social media. It is the first in the literature to discuss the necessity of considering all user engagement indicators and proposes a reliable method that calculates their relative importance.

Citation

Güner, S., Cebeci, H.I. and Aydemir, E. (2023), "How popular is a topic on social media? A multi-criteria decision-making framework based on user engagement", Kybernetes, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/K-05-2023-0884