Answers are grounded in this paper and cite their sources.
Social media represents a vital actor in our lives, often serving as a primary source of information, surpassing traditional sources. Among these platforms, X-Platform which used to be called Twitter has emerged as a leading space for the exchange of opinions and emotions. In this study, we introduce a supervised machine learning system designed to detect user credibility in this influential platform. User’s credibility detection depends largely on the features of this user in the platform. Feature weighting plays a pivotal role in identifying the significance of each feature in a given dataset. It has the ability to indicate irrelevant features which can lead to better performance of classification problems. This paper aims to highlight the impact of weighted features on the accuracy of X-Platform User Credibility Detection (XUCD) using supervised machine learning methods such as Principal Component Analysis (PCA) and Correlation-Coefficient algorithms, and Tree-based methods such as (ExtraTressClarifier) to extract new weighted features in the dataset and then use them to train our model to discover their impact on the accuracy of user credibility detection issue. As a result, we measured the effectiveness of different feature weighting methods on different dataset categories to find out which obtained the best detection accuracy. Experiments are conducted on real user profiles, statistical, and emotional information is extracted from a publicly available dataset called (ArPFN). The improvement of XUCD Accuracy using different weighting methods was dependable on the method and the dataset category that has been used.
No takes yet. Share an insight, caveat, or question.
Abid-Althaqafi et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: