Abstract
A method for selecting potential spam channels for human review by applying a minimum set-cover approximation algorithm to a similarity graph of channels. The nodes of the similarity graph represent channels and the edges between nodes represent similarity between channels. Channels may be reviewed by human reviewers to determine if they are spam. If a channel is connected to a certain number of identified spam channels then they can also be identified as spam channels and suspended. The algorithm may be an iterative process of selecting a channel to review that is connected to the most number of other channels that require another connected spam channel to be suspended.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Moșoi, Alexandru and Noever, Andreas, "SPAM IDENTIFICATION USING SET-COVER ALGORITHM ON A CHANNEL SIMILARITY GRAPH", Technical Disclosure Commons, (August 29, 2019)
https://www.tdcommons.org/dpubs_series/2450