[Paper Review] Identifying Fake Profiles in LinkedIn

Shalinda Adikari, Kaushik Dutta|arXiv (Cornell University)|Jun 2, 2020

Spam and Phishing Detection16 references48 citations

TL;DR

The paper identifies the minimal publicly available LinkedIn profile data necessary to detect fake profiles and achieves 87% accuracy with a 94% true negative rate using a data-mining approach, improving accuracy by ~14% over comparable methods.

ABSTRACT

As organizations increasingly rely on professionally oriented networks such as LinkedIn (the largest such social network) for building business connections, there is increasing value in having one's profile noticed within the network. As this value increases, so does the temptation to misuse the network for unethical purposes. Fake profiles have an adverse effect on the trustworthiness of the network as a whole, and can represent significant costs in time and effort in building a connection based on fake information. Unfortunately, fake profiles are difficult to identify. Approaches have been proposed for some social networks; however, these generally rely on data that are not publicly available for LinkedIn profiles. In this research, we identify the minimal set of profile data necessary for identifying fake profiles in LinkedIn, and propose an appropriate data mining approach for fake profile identification. We demonstrate that, even with limited profile data, our approach can identify fake profiles with 87% accuracy and 94% True Negative Rate, which is comparable to the results obtained based on larger data sets and more expansive profile information. Further, when compared to approaches using similar amounts and types of data, our method provides an improvement of approximately 14% accuracy.

Motivation & Objective

• Motivate the need to detect fake profiles in professionally oriented networks like LinkedIn.
• Determine the minimal set of publicly available LinkedIn profile data required for fake profile identification.
• Develop a data mining approach capable of identifying fake profiles with limited data.
• Evaluate the proposed method against approaches using similar data quantities to assess performance gains.

Proposed method

• Identify and restrict to the minimal publicly available LinkedIn profile data necessary for detection.
• Apply a data mining approach for fake profile identification using the selected features.
• Evaluate performance in terms of accuracy and true negative rate (and compare to larger-data baselines).

Experimental results

Research questions

• RQ1 What is the smallest subset of LinkedIn profile data sufficient to identify fake profiles?
• RQ2 How well can fake profiles be detected with limited publicly available data?
• RQ3 How does the proposed method compare to other approaches using similar data in terms of accuracy and false positives?

Key findings

• With limited profile data, the method achieves 87% accuracy.
• True Negative Rate reaches 94%.
• The approach outperforms comparable methods using similar amounts and types of data by about 14% in accuracy.