GIMENEZ, P. J. A. ; SIQUEIRA, S. W. M. . How much do I Stand Out in Communities Q&A? An Analysis of User Interactions based on Graph Embedding. In: SBSI 2021: XVII Brazilian Symposium on Information Systems, 2021, Uberlândia. SBSI 2021: XVII Brazilian Symposium on Information Systems. Porto Alegre: SBC, 2021. p. 1-8. doi: 10.1145/3466933.3466966
How much do I Stand Out in Communities Q&A? An Analysis of User Interactions based on Graph Embedding
Authors
Paulo Jose de Alcantara Gimenez (UNIRIO)
Sean Wolfgand Matsui Siqueira (UNIRIO)
Abstract
The interactions in Communities Question Answer (CQA) have high dimensionality, generating dispersed and vast information about the users’ behavior. Understanding this behavior and what characteristics qualify users as the best contributors is still a challenge. In this paper, we rely on Persuasion Theory to identify users who stand out at CQA. We use graph incorporation for reducing data dimensionality to analyze six communities. As a result of the experiments, we found a strong correlation between the intensity of user activity and reputation, but it is not a linear relationship. Also, contrary to the literature, the best contributors are not the top 10-20%, but it varies from community to community. With the results of this work, users can design better strategies for collaboration; headhunters can identify the best talents; marketing companies can identify influencers, and developers can adapt their social reputation algorithms.
Keywords:
CQA, Q&A, Social network analysis, User classification, Feature learning
