A community scale test for node affiliation based on network sampling and wavelet analysis.

This paper addresses the challenge of determining community scale to which network nodes belong and introduces an innovative hypothesis testing approach. It begins with a network sampling method that generates sequences of node dependency values, revealing community structure and scale in a waveform-like manner. Subsequently, the study employs wavelet analysis, a signal processing technique, to extract local signal periodicity information from these sequences. This information is then used to develop a test for assessing node membership in specific community scales. The proposed method is applied to both simulated and real-world social network data, with results from the simulated data demonstrating its effectiveness in evaluating node membership in particular community scales. [ABSTRACT FROM AUTHOR]