Here are some complicated solutions that my team members have found to counteract the issues of spreading misinformation:
Imeter based algorithm- Its purpose is to Exploit a point of origin to neighboring nodes in the area to estimate where the attack has originated from. These nodes are intersections points of data which are passed along to other nodes. So when a breach occurs within a node it affects the following nodes surrounding it. So to detect where the attacker might have breached, they use an I-meter and Reverse diffusion to get the exact estimate to where the attack occurred.
Reverse diffusion can deduct which node was attacked by following a trail back to each hitmarker until you are back to the original source, but it’s more complex dealing with a web of nodes. In this case the I-meter can count the amount of hits on each node using the process of reverse diffusion, depending on how many hits on the node, the higher the count is the more possible it is that it’s an attacker.
Rumor Detection Model(CNT) Proposed by Qazvinian et al. – It adopts a variety of features such as content-based features (e.g., words and segments appearance, part of speech), network-based features (i.e., re-tweets or tweets propagation) and twitter-specific memes (i.e., Hashtag or shared URLs). CNT orchestrates an array of strategies to select features to detect misinformation in micro blogs.
Fake News Detector Applications Some of these include:
B.S Detector – Alerts users of unreliable news sources by searching all links of a given web page for sources that have been collected in a unreliable-news database.
PolitiFact – is a six-dimensional rating system developed to check facts. It is frequently used to rate the accuracy and credibility of claims made by US officials and others.
Fake News Detector AI – identifies fake-news websites by measuring similarity to existing fake-news websites using artificial intelligence techniques as a blackbox.
