Visualizing norm polarities for ethical filtering using color-encoded normative tag cloud

Online content has been recently subjected to ethical filtering in conformance to regional set of norms. Normative judgments are often difficult to determine the limits to dispersion and free-flow of information by state and society. Filtering unethical website content requires keyword analysis, a challenge in current manual IP filtering technologies. Due to sparse representation of objectionable words in online content, the seed sets need to focus on positive norms in order to segregate norm violations. In this research work, we exploit these polarities in norm semantics to solve the problem of ethical filtering. We scan text for belief-conforming ethical norms, and filter out words that fall below the norm threshold. Using the Noahide Norms mapped to seven colors and conforming scores, we visualize normative tag cloud. The objective of our positively oriented normative tag cloud is to conform and filter raw text according to the ethical dimensions. Our method has a high classification accuracy compared to manual gold standard. Hence, it can accurately summarize and illustrate visual comparison and contrast of cultural norms out of plain English text.