Trusted News Network - project based on Aigents platform

in #news6 years ago

Based on the Aigents platform (https://aigents.com/), we propose to create a solution for personalized verification and actualization of the news on the Internet, including social networks and online publications, based on the social-evidence based model of consumption and spread of information implemented with methods of experiential machine learning on large amounts of data.
aigents_news.png

The goal of the project is the following:

  • solving the problem of “viral” spread of false news (“fake news”) and distorted interpretation of facts, including biased information (“partisanship”) and “fake” publications on behalf of authoritative authors sources;
  • solving the problem of “social chain reaction” in the post-information society, when limited amount of unreliable or distorted information quickly spread in the cyber space can lead to disastrous consequences in the real world;
  • creating an opportunity for a wide user audience (free of charge), as well as a narrow professional audience (for a fee) to receive verified information from selected sources verified by a selected expert community;
  • creating an opportunity for the expert community to earn both on the examination and verification of materials, and high-quality author publications, confirmed by relevant expert assessments.


Opportunities and prospects of the Aigents platform for social analytics, assessment of news relevance and detection of “fake news” (video and presentation slides)
http://aigents.com/papers/2017/Aigents-News-Analytics-2017-en.pdf

The solution can be created in one of two versions of the product, or both:

  • B2C — Software-as-a-Service (SaaS) cloud-based web application a solution in the form of an SAAS service for end users (individual or corporate) for several main software platforms with a custom graphical or interactive (chat-bot) interface;
  • B2B — “box” solution in one of the two forms:
    corporate web application installed in the corporate network for employees of enterprises or as a news platform;
    new syndication portal, specific to vertical market or region, nation or community, deployed by respective vendors for local markets.

Within the project, it is proposed to implement the following functions for the B2C and B2B markets:

  • automatic aggregation of news by sources and topics specified by both the users themselves and the expert community they have selected (partially implemented and available now);
  • automatic determination of the subject matter, relevance and tonality according to formal criteria determined at the expert level as well as based on collective intelligence by means of explicit and implicit peer review, taking the preferred expert user community into account (partially implemented and available now);
  • automatic formation of a database of formal criteria for the tasks indicated above, relying on machine learning based on expert and user assessments and data;
  • formation of ratings of news sources, authors and topics on the basis of assessments of the news they supply – specific to sections of selected expert communities;
  • generation of income for authors of publications and experts who provide assessments and determine the criteria of relevance and reliability – depending on the totality of estimates confirmed by paid users;
  • identification and blocking distributors of unreliable and destructive content accordingly to community-specific regulations – including automated “bots” and personified “trolls”.

The implementation of the solution to be based on the following principles:

  • social-evidence based model of consumption and spread of information, implemented with methods of experiential machine learning on large amounts of data;
  • multi-level authentication of users, experts and authors based on e-mail, telephone number, social network connections, decentralized identification systems (http://openid.net/) and local state identification services;
  • financial settlements between users and authors based on the widespread payment solution – such as an electronic payment system, a large bank, a blockchain or a combination of them, moreover, the company providing the solution can act as a general partner of the project.


Cognitive architecture of collective intelligence based on social evidence (video and paper)
http://www.math.nsc.ru/AP/ScientificDiscovery/PDF/kolonin_cognitive_architecture.pdf

Project status at the moment:

  • there is a prototype web application are the site https://aigents.com/, partially solving the stated problems;
  • integration of different sources information for search and monitoring such as any web sites with static html and social networks such as Facebook, Google+, VKontakte and blockchains such a Ethereum, Steemit and Golos;
  • ongoing public beta-testing of the solution with initial user base in the amount of 600+ users and about 30 monthly users;
  • the project requires expert competences in the field of large international business in the field of IT and mass media – to develop a business plan and a monetization scheme, build an international business and bring the product to the market;
  • depending on the decisions made regarding the business plan, the project may require a financial settlement partner (a large bank with an international reputation, a payment system or a blockchain).


Targeted pro-active information search with adaptive intelligent Internet agents (video)

The following monetization options are possible:

  • B2C – commissions from the "author's reward" granted to authors and experts, redistributed based on the subscription price of paid users of the cloud-based web application monthly;
  • B2B – corporate licenses for the installation and use of a “box” solution by corporate or local (in highly specialized, regional and national markets) consumers, as well as support and after-sales service.