[Paper Review] Predicting Social Status via Social Networks: A Case Study on University, Occupation, and Region

Hao Fu, Xing Xie+4 more|arXiv (Cornell University)|Apr 10, 2016

Complex Network Analysis Techniques22 references3 citations

TL;DR

This paper proposes a method to predict real-world social status—specifically university prestige, occupation status, and regional development—using social network structure from Sina Weibo. By modeling user-group memberships as a heterogeneous network and applying a supervised random walk-based inference algorithm, the authors achieve reasonable accuracy in predicting group-level social status, though inconsistencies arise due to user bias and online behavior manipulation.

ABSTRACT

Social status refers to the relative position within the society. It is an important notion in sociology and related research. The problem of measuring social status has been studied for many years. Various indicators are proposed to assess social status of individuals, including educational attainment, occupation, and income/wealth. However, these indicators are sometimes difficult to collect or measure. We investigate social networks for alternative measures of social status. Online activities expose certain traits of users in the real world. We are interested in how these activities are related to social status, and how social status can be predicted with social network data. To the best of our knowledge, this is the first study on connecting online activities with social status in reality. In particular, we focus on the network structure of microblogs in this study. A user following another implies some kind of status. We cast the predicted social status of users to the "status" of real-world entities, e.g., universities, occupations, and regions, so that we can compare and validate predicted results with facts in the real world. We propose an efficient algorithm for this task and evaluate it on a dataset consisting of 3.4 million users from Sina Weibo. The result shows that it is possible to predict social status with reasonable accuracy using social network data. We also point out challenges and limitations of this approach, e.g., inconsistence between online popularity and real-world status for certain users. Our findings provide insights on analyzing online social status and future designs of ranking schemes for social networks.

Motivation & Objective

• To investigate whether social network structure can serve as a proxy for real-world social status, including university prestige, occupation, and regional development.
• To address the challenge of inferring social status from incomplete and biased user-provided profile data in large-scale online social networks.
• To develop a scalable and effective method for predicting group-level social status using network structure and membership information.
• To validate predicted status scores against real-world indicators such as institutional prestige and occupational rankings.
• To identify limitations and biases in online social status prediction, particularly related to user self-presentation and network behavior manipulation.

Proposed method

• Model user-group membership (e.g., university, occupation, region) as a heterogeneous network with two node types: users and groups.
• Formulate group status prediction as a ranking problem in the heterogeneous network, where group status is a real-valued measure reflecting member social status.
• Apply a supervised random walk algorithm to learn edge weights and propagate status scores from known memberships to infer missing attributes.
• Use known user attributes (e.g., university, occupation) as training signals to guide the random walk process and improve inference accuracy.
• Aggregate predicted user statuses to compute group-level status scores, which are then compared with real-world benchmarks.
• Leverage network structure and user behavior patterns to infer social status without relying on explicit content or textual analysis.

Experimental results

Research questions

• RQ1 Can social network structure accurately predict real-world social status indicators such as university prestige, occupational status, and regional development?
• RQ2 How does the accuracy of social status prediction vary across different domains (university, occupation, region) when using network-based inference?
• RQ3 To what extent do biases in user-provided profile data and online behavior (e.g., link farming, self-promotion) distort the prediction of social status?
• RQ4 How effective is the supervised random walk approach in handling incomplete and noisy membership data in large-scale social networks?
• RQ5 What are the key limitations of using online popularity as a proxy for real-world social status, and how can they be mitigated?

Key findings

• The proposed method achieves reasonable accuracy in predicting social status for universities, occupations, and regions using only social network structure and partial membership data.
• A significant inconsistency is observed between online popularity and real-world social status, particularly for users in creative or self-promotional professions such as artists and marketers.
• Elite graduates from non-top universities are more likely to have verified accounts, indicating a bias in user self-presentation that skews network-based status measures.
• Officials and religious leaders often conceal their identities online, leading to underrepresentation in membership inference and status prediction.
• Users in certain occupations—especially in arts and entertainment—exhibit higher online activity and follower counts, which may inflate their perceived social status relative to their real-world standing.
• The study identifies that link farming and strategic networking practices can artificially elevate online status, highlighting a key challenge for future models to account for such behaviors.