Speaking at the simplest level, viral information is articles, videos and posts on social networks that are completely untrue, i.e. are deceitful. These materials are aimed at people on the Internet who share similar interests, in the hope that they will be reposted. And since people rarely check the sources or the facts behind these articles, they can quickly spread like a virus.
Fake news can misshape people's perceptions and attitudes about something, creating suspicion, mistrust, and causing them to reject factual truths. And very often these fake news sites are infected with malware and viruses to steal your personal data. Currently, social media platforms such as Facebook and Twitter have been heavily criticized for allowing fake news to be spread using their services. But the fact is that their profits are closely related to the people who read and click on the stories shown in their networks. Therefore, there were few incentives for the correct solution of this problem. However, today, identifying fake news is an interdisciplinary task. To solve it, experts in human behavior, mass communications, mathematics and statistics, linguistics, information technology, security and other areas are involved.
The most promising for solving these problems is the use of artificial intelligence in virus information recognition systems. And the result, which may be recognized as effective in the future, will most likely also be based on an integrated approach. Well, now let's talk a little more: At one stage, computational linguistics tools can be used and texts can be classified according to established criteria. On the other hand, metadata is analyzed: how often and at what time the account is active, under what login it is registered, what topics it writes on and what triggers it reacts to. It can also analyze the technical data of the browser and device, as well as other available activity parameters.
Based on the results of the analysis, the system can mark the news as potentially fake, and this information can be sent to an employee of a special service for verification. So, solutions that work according to the described principle already exist, are tested and put into practice: Botometer and BotoSlayer, developed by Indiana University Bloomington (USA), Tealeaf by IBM Sentimetrix, detection systems used by Facebook and other major social networks. However, it was found that such an algorithm made the right decision in 76% of cases. This means that the verification system still makes a relatively large number of errors, but it still outperforms a human at deception detection.
All of the above problems and limitations do not stop the development and development of automatic detection methods and help, if not win, then not lose in the war of facts and fakes and detect fakes almost at the same speed with which they are produced. One of such systems to increase the accuracy of the checked content was created by the Exorde system. Exorde is built around a core platform that provides unbiased credibility scores for information (and virality-related analytics) based on the community, AI modules, and a token-based economy. This platform is decentralized, open and transparent. This is where participants will work together to index the entire network, extract its unstructured information, relationships, similarities, trends, and any type of pattern in information circulating throughout the internet, regardless of platform or media. Exorde as a knowledge network can use the unstructured content of the Internet and perform a first-of-its-kind analysis of the virality of information circulating throughout the network.
The Exorde Work Systems are the core component and will serve as the base layer for the entire ecosystem. Exorde is based on four main parameters: Ethereum is an open source platform that allows developers to create and deploy decentralized applications on it. Skale is an elastic sidechain network connected to the Ethereum chain, acting as an execution layer. It is a scalable and elastic environment that provides high transaction throughput. Filecoin is a decentralized data storage system that intends to replace the expensive and inefficient servers of Google and Amazon. NLP is an artificial intelligence module that will work with unstructured text. AI will be able to recognize objects in the text, extract sentences, facts, compare them with each other and perform similarity checks.
Join to Discord: https://discord.gg/e4y7bjQSwW
#crypto #web3 #exorde @ExordeLabs
https://blog.coinlist.co/introducing-the-coinlist-seed-winter-2022-batch/
https://exorde.network/