Application of the Benford’s law to Social bots and Information Operations activities

No ratings

Presented at CyberScience2020 2020 by

Lale Hall

Benford's Law shows the pattern of behavior in normal systems. It states that in natural systems digits'frequency have a certain pattern such that the frequency of numbers' first digits is not evenly distributed. In systemswith natural behavior, numbers begin with a "1" are more common than numbers beginning with "9". It implies that ifthe distribution of first digits is outside of the expected distribution it can be indicative of fraud. It has many applicationsin forensic accounting, stock markets, finding abnormal data in survey data, and natural science. We investigatewhether social media bots and Information Operations activities are conformant to the Benford’s law. Our resultsshowed that bots’ behavior adhere to Benford's Law, suggesting that using this law helps in detecting maliciousonline automated accounts and their activities on social media. However, activities related to Information Operationsdid not show consistency in regards to Benford’s Law. Our findings shed light on the importance of examining regularand anomalous online behavior to avoid malicious and contaminated content on social media.